Commercial Building Façade Design:

Improving the Consistency of Early Design Tool Predictions

and Detailed Design Tool Calculations

By

Chi-Yao Hsu

A thesis
submitted to the School of Architecture, Victoria University of Wellington,
in fulfilment of the requirements for the degree of
Master of Building Science.

Victoria University of Wellington

July 2011

i
ACKNOWLEDGEMENTS

Firstly, I would like to dedicate this research thesis to one of my best friends, the late
Geoffrey Payne. Rest in peace brother, love you and miss you always. To the Payne family,
thanks for all the great memories throughout the years; Geoffrey will be forever in our hearts.

I would like to mention a special acknowledgement to my family and friends, their support and
encouragement made this all possible. I love you Mum & Dad, thanks for always reminding me
to “provide for a rainy day”. It has been a year to remember, thank you all for believing in me
and being there for me when it mattered the most.

Many thanks to my research supervisor, Michael Donn, who provided great guidance and
encouragement on this research thesis, special thanks for his enthusiasm and valuable
feedback which made this project possible.

Special thanks to all the architects and engineers who provided me with great feedback, it was
fascinating and just simply inspiring.

Big thanks to the professionals at Lawrence Berkeley National Laboratory, Berkeley, California,
USA for their support in the use of COMFEN.

Finally, I would like to say thanks to Nigel Isaacs and BRANZ for the Building Energy End-use
Study (BEES) Masters Scholarship.

i
ii
PREFACE

This research thesis was submitted in fulfilment of the requirements for the degree of Master of
Building Science at the School of Architecture, Victoria University of Wellington, New Zealand.

Author: Chi-Yao Hsu

School of Architecture

Victoria University of Wellington

Email: [email protected]

Contact number: +64 21 101 3243

Research Supervisor: Michael Donn

Director of Centre for Building Performance Research

School of Architecture

Victoria University of Wellington

Email: [email protected]

Contact numbers:
+64 4 463 6221 work
+64 4 463 6204 work fax
+64 21 611 280 mobile
+64 21 611 594 mobile fax

Postal address: PO Box 600 Wellington, New Zealand

iii
iv
Table of Contents

ACKNOWLEDGEMENTS ............................................................................................................... i

PREFACE ..................................................................................................................................... iii

List of Figures ................................................................................................................................ ix

List of Tables ................................................................................................................................. xi

ABSTRACT ................................................................................................................................... 1

Key Definitions & Abbreviations .................................................................................................... 3

1 INTRODUCTION........................................................................................................................ 5

1.1 Focus: Performance Sketch Concept ................................................................................. 6

1.2 Aim & Objectives ................................................................................................................. 8

1.3 Lessons from The Architectural Sketch .............................................................................. 9

1.4 Research Overview ........................................................................................................... 14

2 LITERATURE REVIEW ............................................................................................................ 15

2.1 Users of the Early Design Tool ......................................................................................... 15

2.2 Relevance of the Early Design Tool.................................................................................. 20

2.3 The Relationship between Early Design Lessons and Detailed Design Lessons ............ 25

2.4 Commercial Building Façade Design ................................................................................ 29

2.5 Integration of Aesthetics & Performance .......................................................................... 36

3 METHODOLOGY ..................................................................................................................... 39

3.1 Overall Hypotheses ........................................................................................................... 39

3.2 Methodology Overview...................................................................................................... 40

3.3 An Example of a Performance Sketch Early Design Tool: COMFEN 3.0 ......................... 41

3.4 An Example of a Detailed Design Tool: EnergyPlus 4.0................................................... 45

3.4.1 OpenStudio Plug-in .................................................................................................... 47

3.5 Introduction of a Performance Sketch Tool to Practitioners ............................................. 48

3.5.1 Interview Sample Size................................................................................................ 49

3.5.2 Delphi-like Review of Responses .............................................................................. 49

3.5.3 Process of Interviewing Practitioners: Introduction of COMFEN ............................... 50

3.6 Consistency Test of a Performance Sketch Tool .............................................................. 51

3.6.1 The Definition of Consistency .................................................................................... 52

3.6.2 Consistency Test: Fenestration Design Scenarios .................................................... 53

3.6.3 Fenestration Design Scenarios Integrated into DoE Benchmark Models .................. 55

v
 3.6.4 Simulation Settings ..................................................................................................... 58

3.7 Evaluating the Adequacy of the Sketch within a Performance Sketch Tool...................... 59

3.7.1 Adding Complexity to the Performance Sketch Tool .................................................. 59

3.7.2 Individual Zone Comparison: Results of Adding Complexity to the Performance

Sketch Tool .......................................................................................................................... 63

3.7.3 A Consistency Test for the Study of Adding Complexity to the Performance Sketch
Tool ...................................................................................................................................... 64

4 RESULTS ................................................................................................................................. 67

4.1 Use-cases: Introduction of a Performance Sketch Tool to Practitioners ........................... 67

4.1.1 Interview Questions .................................................................................................... 67

4.1.2 Overall Analysis of Conclusions: Use-cases .............................................................. 68

4.2 Consistency Test of a Performance Sketch Tool .............................................................. 72

4.2.1 Exploration of Possible Climatic Differences .............................................................. 72

4.3 Evaluating the Adequacy of the Sketch within a Performance Sketch Tool...................... 79

4.3.1 Adding Complexity to the Performance Sketch Tool .................................................. 79

4.3.2 Individual Zone Comparison: Results of Adding Complexity to the Performance

Sketch Tool .......................................................................................................................... 83

4.3.3 A Consistency Test for the Study of Adding Complexity to the Performance Sketch
Tool ...................................................................................................................................... 86

5 Practitioner-based Performance Sketch Tool Specifications ................................................... 89

5.1 Freedom of Aesthetics ....................................................................................................... 90

5.2 Sustainable Designs .......................................................................................................... 91

5.3 Delivering the Design......................................................................................................... 92

5.4 Diagram of Decision-making for Performance Sketch Analysis ........................................ 93

6 CONCLUSIONS ....................................................................................................................... 95

FUTURE WORK .......................................................................................................................... 97

REFERENCES ............................................................................................................................ 99

APPENDICES ............................................................................................................................ 105

APPENDIX A – Selection of Practitioners: Information Sheet............................................... 105

APPENDIX B – Selection of Practitioners: Consent Form .................................................... 106

APPENDIX C – Practitioner A: Initial Interview Feedback .................................................... 107

APPENDIX C.1 – Practitioner A: Analysis of Initial Interview Feedback ............................... 108

nd
APPENDIX C.2 – Practitioner A: 2 Round of Feedback ..................................................... 109
nd
APPENDIX C.3 – Practitioner A: Analysis of 2 Round of Feedback .................................. 111
vi
APPENDIX D – Practitioner B: Initial Interview Feedback .................................................... 112

APPENDIX D.1 – Practitioner B: Analysis of Initial Interview Feedback .............................. 113

APPENDIX E – Practitioner C: Initial Interview Feedback .................................................... 114

APPENDIX E.1 – Practitioner C: Analysis of Initial Interview Feedback .............................. 115

APPENDIX F – Practitioner D: Initial Interview Feedback .................................................... 116

APPENDIX F.1 – Practitioner D: Analysis of Initial Interview Feedback .............................. 117

APPENDIX G – Practitioner E: Initial Interview Feedback .................................................... 118

APPENDIX G.1 – Practitioner E: Analysis of Initial Interview Feedback .............................. 119

APPENDIX H – Practitioner F: Initial Interview Feedback .................................................... 120

APPENDIX H.1 – Practitioner F: Analysis of Initial Interview Feedback .............................. 121

APPENDIX I – Façade Constructions & Material Properties ................................................ 122

APPENDIX J – Simulation Location: Wellington, New Zealand ........................................... 123

APPENDIX J.1 – Simulation Location: Wellington, New Zealand ........................................ 124

APPENDIX K – Simulation Location: Melbourne, Australia .................................................. 126

APPENDIX K.1 – Simulation Location: Melbourne, Australia ............................................... 127

APPENDIX L – Simulation Location: Minneapolis-St Paul, USA .......................................... 128

APPENDIX L.1 – Simulation Location: Minneapolis-St Paul, USA ....................................... 129

APPENDIX M – Simulation Location: Bombay, India ........................................................... 130

APPENDIX M.1 – Simulation Location: Bombay, India ........................................................ 131

vii
viii
List of Figures

Figure 1 – Sketch of the New York Times building ....................................................................... 9

Figure 2 – Exterior render of the New York Times building in New York, USA ............................ 9

Figure 3 – Sketch of the UNIQA-Tower ...................................................................................... 10

Figure 4 – Exterior view of the UNIQA-Tower in Vienna, Austria ............................................... 10

Figure 5 – Sketch of the Sentinel building .................................................................................. 10

Figure 6 – Exterior view of the Sentinel building in Glasgow, UK ............................................... 10

Figure 7 – Sketch of the Tech Gate Tower ................................................................................. 11

Figure 8 – Exterior view of the Tech Gate Tower in Vienna, Austria .......................................... 11

Figure 9 – Sketch of the Torre Almirante building....................................................................... 11

Figure 10 – Exterior view of the Torre Almirante building in Rio de Janeiro, Brazil.................... 11

Figure 11 – Sketch of the Koniglich-Niederlandische Botschaft building ................................... 12

Figure 12 – Exterior view of the Koniglich-Niederlandische Botschaft in Berlin, Germany ........ 12

Figure 13 – Sketch of the Spherion-Office building Deloitte ....................................................... 12

Figure 14 – Exterior view of the Spherion-Office building Deloitte in Dusseldorf, Germany ...... 12

Figure 15 – Workflow and subtasks in performance-based design (Kalay, 1999) ..................... 28

Figure 16 – Exterior view of Taipei 101 in Taipei, Taiwan .......................................................... 32

Figure 17 – Exterior view of the Audi Forum building in Tokyo, Japan ...................................... 32

Figure 18 – International Finance Centre in Hong Kong. Image © (Crosbie, 2005) ................... 33

Figure 19 – Exterior view of the Mode Gakuen Cocoon Tower in Tokyo, Japan ....................... 37

Figure 20 – Exterior view of the Mode Gakuen Spiral Towers in Nagoya-shi, Japan ................ 37

Figure 21 – COMFEN 3.0 simulation interface ........................................................................... 41

Figure 22 – COMFEN fenestration editing view in elevation ...................................................... 43

Figure 23 – COMFEN fenestration editing view in plan .............................................................. 43

Figure 24 – COMFEN fenestration editing view in section ......................................................... 43

Figure 25 – EnergyPlus 4.0 idf-Editor ......................................................................................... 45

Figure 26 – OpenStudio within SketchUp ................................................................................... 47

Figure 27 – Façade designs simulated in COMFEN & EnergyPlus............................................ 53

Figure 28 – Typical COMFEN single-zone model and its dimensions ....................................... 54

Figure 29 – Geometry of DoE small benchmark office ............................................................... 55

Figure 30 – Geometry of DoE large benchmark office ............................................................... 55

ix
Figure 31 – Floor layout of perimeter zones and core zone in the small benchmark office ........ 56

Figure 32 – Floor layout of perimeter zones and core zone in the large benchmark office ........ 56

Figure 33 – Plan view: modelling categories ............................................................................... 60

Figure 34 – Façade designs simulated in consistency test ......................................................... 64

Figure 35 – COMFEN and EP simulations in Melbourne ............................................................ 74

Figure 36 – COMFEN and EP simulations in Bombay ................................................................ 74

Figure 37 – COMFEN and EP simulations in Wellington ............................................................ 75

Figure 38 – COMFEN and EP simulations in Minneapolis .......................................................... 75

Figure 40 – Mixed Mode with daylight control simulations .......................................................... 81

Figure 41 – Consistency test: Mixed Mode models with daylight control .................................... 86

Figure 42 – COMFEN perimeter zones with an additional generic core zone ............................ 92

Figure 43 – Diagram of decision-making for performance sketch analysis ................................. 93

x
List of Tables

Table 1 – Floor area percentage of perimeter zones versus core zones ................................... 57
Table 2 – General simulation settings used in the consistent test of COMFEN ......................... 58
Table 3 – Further detailed simulation settings in COMFEN and DoE offices ............................. 58
Table 4 – General simulation settings for 1~3-zone models ...................................................... 62
Table 5 – Detailed simulation settings for 1~3-zone models ...................................................... 62
Table 6 – Advantages of COMFEN ............................................................................................ 68
Table 7 – Desired design features in future versions ................................................................. 69
Table 8 – Consistency of COMFEN fenestration ranking ........................................................... 72
Table 9 – Annual energy distribution of designs simulated in Bombay ...................................... 77
Table 10 – Annual energy distribution of designs simulated in Wellington ................................. 77
Table 11 – Annual energy distribution of HVAC Only models simulated in Wellington .............. 80
Table 12 – Annual energy distribution of Mixed Mode models simulated in Wellington ............. 81
Table 13 – Distribution of energy use for each zone within the COMFEN model ...................... 83
Table 14 – Distribution of energy use for each zone within the 2-zone model ........................... 83
Table 15 – Distribution of energy use for each zone within the 3-zone model ........................... 83
Table 16 – EUI percentage comparison of the modelling categories simulated in Wellington ... 84
Table 17 – Consistency test: Mixed Mode models with daylight control .................................... 87

J. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Wellington .......................... 123

J. 2 – COMFEN Double clear Low-E between VB 45 simulated in Wellington ........................ 123
J. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Wellington ........................... 123
J. 4 – COMFEN Single clear simulated in Wellington ............................................................... 123
J. 1. 1 – Annual energy distribution of designs simulated in Wellington ................................... 124

K. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Melbourne ......................... 126

K. 2 – COMFEN Double clear Low-E between VB 45 simulated in Melbourne ........................ 126
K. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Melbourne .......................... 126
K. 4 – COMFEN Single clear simulated in Melbourne .............................................................. 126
K. 1. 1 – Annual energy distribution of designs simulated in Melbourne .................................. 127

L. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Minneapolis ....................... 128

L. 2 – COMFEN Double clear Low-E between VB 45 simulated in Minneapolis ...................... 128
L. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Minneapolis ........................ 128
L. 4 – COMFEN Single clear simulated in Minneapolis ............................................................ 128
L. 1. 1 – Annual energy distribution of designs simulated in Minneapolis ................................ 129

M. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Bombay ............................ 130

M. 2 – COMFEN Double clear Low-E between VB 45 simulated in Bombay ........................... 130
M. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Bombay ............................. 130
M. 4 – COMFEN Single clear simulated in Bombay ................................................................. 130
M. 1. 1 – Annual energy distribution of designs simulated in Bombay ..................................... 131

xi
xii
ABSTRACT

The focus of this research is the concept of the 'Performance Sketch' tool. This is to use
detailed simulation software to calculate (plausible) energy performance of designs quickly.
Analogous to the Architectural Sketch the Performance Sketch uses high quality tools (detailed
simulation) to create an accurate, but simple representation of the essential properties of a
building, as opposed to a detailed representation.

The aim of this research is to assess the consistency between the predictions produced by
performance sketch design tools and the calculations produced by detailed design tools.

The Lawrence Berkeley National Laboratory’s (LBNL) computer software COMFEN

(COMmercial FENestration) is a performance sketch tool. It makes the power of the complex
detailed design simulation package EnergyPlus available in the very early stages of the design
process. It uses a single zone, single external façade EnergyPlus model to explore the costs
and benefits of alternate façade designs.

The hypothesis tested is that the COMFEN (single-zone) energy performance calculation
method is plausible for early design analyses. It evaluates the performance sketch approach
from three different points of view: first, COMFEN was introduced to various practitioners in the
building industry to gather use-case feedback on the performance sketch approach. A list of
specifications for performance sketch design tools was developed based on these use-cases.
Second, it examines whether the optimum façade identified by COMFEN creates the optimum
performance complex building when this optimum façade is incorporated into detailed building
models. Finally, refinements of the nature of the performance sketch based on this use-case
feedback were tested in EnergyPlus.

The thesis concludes by drawing together these three threads into an outline of a practitioner-
based definition of an ideal performance sketch which has been tested in practical application.

1
2
Key Definitions & Abbreviations

Definitions

COMFEN: An example of a Performance Sketch early design tool used for this research
(single-zone simulations); it uses EnergyPlus (a detailed design simulation tool) to quickly
calculate energy performance of building façade designs.

Consistency: COMFEN/single-zone and EnergyPlus/multi-zone predictions of energy

performance will produce the same design advice/feedback (i.e. the ranking of the façade
ideas will be the same, as well as similarity in distribution of energy end-uses).

Detailed design lessons: The lessons learnt from the detailed design tool.

Detailed Design Tool (DDT): A DDT is a complex building energy simulation tool
(e.g. EnergyPlus) used during the detailed design process, where all the details are included
in the computer model.

Early design lessons: The lessons learnt from the early design tool.

Early Design Tool (EDT): An EDT is a building energy simulation tool used during the early
design process, where a small amount of detail is available.

EnergyPlus (EP): An example of a detailed design tool used for this research (multi-zone
simulations).

Multi-zone: Complex computer simulations in EnergyPlus (time consuming calculations);

modelling of associated spaces within a building design.

Performance Sketch Tool: It uses a detailed design tool on a sketch of the building design.
For example, COMFEN allows the user to sketch building designs with the power of a
detailed simulation tool (e.g. EnergyPlus).

Single-zone: Simple 1 zone computer simulations in COMFEN (quick calculations).

Use-cases: Feedback based on the introduction of COMFEN program to practitioners.

Abbreviations

ASHRAE: American Society of Heating, Refrigerating, and Air-Conditioning Engineers

COMFEN: COMmercial FENestration

DoE: Department of Energy (USA)

EUI: Energy Use Index

HVAC: Heating, Ventilation, and Air-Conditioning

IWEC: The International Weather for Energy Calculation

LBNL: Lawrence Berkeley National Laboratory (Berkeley, CA, USA)

NREL: National Renewable Energy Laboratory (USA)

TMY: Typical Meteorological Year

3
4
1 INTRODUCTION

This is a usability study focusing on the ‘Performance Sketch’. The concept of the performance
sketch is to use a detailed simulation tool that calculates energy performance of designs quickly
because it: uses a sketch model that is simple and easy to understand; is an accurate ‘sketch’
of the important environmental features; and avoids the complexity of interoperability with CAD
software in order to gain speed and efficiency.

Early design tools play a crucial role in a building project. To design an energy efficient building,
the design team must simulate the mechanical system from day one of the project (i.e.
determine the size of equipment); and carefully assess comfort and performance issues. Like
the architect’s sketch, if it is as flexible in application as the ideal performance sketch, then the
early design tool can be returned to at any stage in the design process. Should it become
necessary to explore a new idea for a façade during developed design, the sketch tool may
allow a rapid analysis of new design options that can be returned to the comprehensive model
once resolved.

The research uses a three-method approach to evaluate the performance sketch concept:

Method 1: An example of a performance sketch early design tool was introduced to various
practitioners for usability feedback. The feedback gathered from interviews around the world
was used to analyse the consistency of performance sketch tools’ energy predictions with the
calculations of detailed design tools. In this research the interview feedback are ‘Use-cases’. “A
use-case defines a sequence of actions a system performs that yields an observable result of
value to a particular actor. A use-case describes primarily functional but also non-functional
requirements from the perspective of an actor achieving particular goals.” (Haumer 2004). The
goal was to work alongside practitioners (architects & engineers) to obtain the opinions of
professionals who would use tools similar to a performance sketch tool on a day-to-day basis.
The significance of this research is that the outcome should help improve the usability and
capability of future performance sketch tools.

Method 2: A performance sketch tool and a detailed design tool were used to examine the
energy performance of various façade designs. The research tests high performance heavily
glazed façades for commercial buildings to examine the relationship/consistency between early
design lessons (single-zone simulations) and detailed design lessons (multi-zone simulations).

Method 3: The feasibility of adding complexity to the performance sketch tool was explored in
order to address some of the concerns of the use-case interviewees (i.e. use-case suggestions:
desired design features). A detailed design tool was used with various levels of model
complexity to explore the balance between simplicity and speed on the one hand and
complexity and lack of response on the other.

This unique three-method approach offers the potential to critique the performance sketch
concept in accuracy and usability terms.
5
1.1 Focus: Performance Sketch Concept

The performance sketch concept is to use a detailed design tool on a sketch of the building
design. It allows the user to sketch building designs with the power of a comprehensive
simulation tool.

COMFEN was used as an example tool for this research because it has characteristics and
properties which match the performance sketch concept:
• It uses a sophisticated and flexible energy modelling tool (e.g. EnergyPlus) but reduces
the complexity of what it is modelling, which provides gains in speed and efficiency
during simulations
• It therefore calculates energy performance of building façade designs quickly
• It is simple and easy to understand
• It is not an exact copy of the appearance of the final building but it is a ‘sketch’ of
the important environmental features of the fenestration system.

COMFEN is a simplified interface to a sophisticated and flexible simulation tool EnergyPlus

(LBNL 2010). It is intended to make it easy to assess fenestration designs for commercial
buildings. This interface has a very different set of input parameters and appearance to stand
alone EnergyPlus. Each interface is intended for a different stage of design (i.e. early or
detailed). With a detailed simulation engine like EnergyPlus the user can examine all essential
design variables (hour by hour). With a Performance Sketch interface like COMFEN, there is
less flexibility, but the calculation has the same rigour.

The inspiration for the research was to explore a series of use-cases and determine
practitioners’ simulation needs from an early design perspective. Essentially, these use-cases
are professionals’ evaluations of the performance sketch concept. Any performance sketch tool
is a simplified model of reality and thus restricts some designs from being simulated
(limitations/parameters). The use-case studies were designed to establish what practitioners
need in a performance sketch interface; so it is reliable and comprehensive enough to kick-start
their project on a solid foundation. The long term goal is to have an early design simulation
package that examines all essential design variables (environmental factors) during the early
stages of design.

Decisions made during the early design stage of a building affect ones made in the detailed
design stage, which ultimately determines project success or failure (Hari 2001).

The hypothesis tested is that by testing the use-cases performance sketch based decisions will
be more relevant (consistent) to multi-zone simulations. In other words, the performance sketch
concept is plausible for early design analyses. Examples of the issues examined are: if the
entire building was modelled during the early design process would the façade design decisions
differ to ones suggested by a performance sketch tool? A further test was whether practitioners
building a performance sketch should be modelling the whole building using multiple zones from
day one?
6
There are two different approaches to the development of an early design tool:
• Explore a range of building design ideas in detail (with simulation) and then
summarise these in a book/tables of data/simple rule of thumb/another simplified
software package – guidance thus derived is dependent on scenarios explored at
the outset.
• Use a sophisticated and flexible modelling tool (e.g. EnergyPlus) but reduce the
complexity of what it is modelling (e.g. COMFEN).

Energy simulation tools are essential in order to forecast a proposed building’s thermal
performance. However, the virtual representation of a building design can sometimes differ
drastically from reality. Mazzarella & Pasini (2009) say:
“Energy savings in buildings is today mandatory in developed countries, so it is
imperative to perform, during the design stage, an accurate estimation of the
energy used by buildings to assure different kinds of comfort.”

7
1.2 Aim & Objectives

The aim of this research was to test whether the performance sketch can provide relevant
advice. The objective was to compare the consistency of energy performance predictions
derived from COMFEN with those from EnergyPlus. This was supplemented by use-case
analyses of the value and relevance of COMFEN as an example of a performance sketch tool.

Ultimately the outcome of this research was to be a specification (based on the use-cases) of
an ideal performance sketch tool supplemented by a decision-making diagram locating the
performance sketch within practitioners’ work flows.

The following questions are examined in this research:

1. What are practitioners’ needs from a performance sketch tool?

E.g. software capabilities and usability; graphical information; relevance of the tool’s
design predictions
2. How relevant is the information predicted by a performance sketch tool?
E.g. accuracy and consistency of design predictions; lessons learnt from these
design predictions
3. Performance Sketch Early Design Tool – how simple is too simple?
E.g. ultimately define whether the design information within a performance sketch
tool is too simple and its interface needs to be more comprehensive
4. How close is close enough?
E.g. difference between early design lessons and detailed design lessons is
significant especially when the design advice/feedback changes

Note: These questions will be addressed and analysed in the Conclusions chapter.

8
1.3 Lessons from The Architectural Sketch

Architectural ideas are expressed through simple and quick sketches (by architects) which
represent the aesthetics of the building. The performance sketch tool has a similar basis:
representing the energy concept of the building.

Figures 1~14 show examples of concept sketches and the completed buildings. These show
that a sketch idea by an architect is expressed through particular principles which communicate
with the client and developer (e.g. shape, scale & context). Practitioners transferred the design
idea established during the early design stage into the physical built form. It is evident that there
is a strong aesthetic relationship between the sketch and actual building. The designers carried
the idea throughout the project. The early design decisions here clearly have had a huge
influence on the final outcome.

Figure 1 – Sketch of the New York Times building

Figure 2 – Exterior render of the New York Times building in New York, USA
Images © Renzo Piano Building Workshop (http://newyorktimesbuilding.com)

9
Figure 3 – Sketch of the UNIQA-Tower
Figure 4 – Exterior view of the UNIQA-Tower in Vienna, Austria
Images © Neumann + Partner, Architekten (Hindrichs & Heusler, 2010)

Figure 5 – Sketch of the Sentinel building

Figure 6 – Exterior view of the Sentinel building in Glasgow, UK
Images © Gordon Murray & Alan Dunlop, Architects (Hindrichs & Heusler, 2010)

10
Figure 7 – Sketch of the Tech Gate Tower
Figure 8 – Exterior view of the Tech Gate Tower in Vienna, Austria
Images © Holzbauer und Partner Architekten (Hindrichs & Heusler, 2010)

Figure 9 – Sketch of the Torre Almirante building

Figure 10 – Exterior view of the Torre Almirante building in Rio de Janeiro, Brazil
Images © Robert A. M. Stern (Hindrichs & Heusler, 2010)

11
Figure 11 – Sketch of the Koniglich-Niederlandische Botschaft building
Figure 12 – Exterior view of the Koniglich-Niederlandische Botschaft in Berlin, Germany
Images © Office for Metropolitan Architecture (Hindrichs & Heusler, 2010)

Figure 13 – Sketch of the Spherion-Office building Deloitte

Figure 14 – Exterior view of the Spherion-Office building Deloitte in Dusseldorf, Germany
Images © Deilmann Koch, Architekten Stadtplaner (Hindrichs & Heusler, 2010)

12
Figures 1~14 show that the design lessons drawn from these sketches (simple ideas before
studied in detail) have been carried out to completion. They suggest once a conceptual (sketch)
idea is chosen it is apparent in the final product (i.e. completed building). This suggests why the
performance sketch analogy may be useful. What is being sketched during the very early
stages of design has a significant impact on a building project.

Early energy prediction is very much like the sketch of a building design (e.g. Figure 1). The
‘simple lines’ within a sketch are comparable to the performance sketch concept used to derive
a theoretical idea of how the building design performs; by analogy, the performance sketch
examines the environmental ideas in a building design. Like the architectural sketch, it must be
simple and quick.

Generally, in early design the best concepts (aesthetic or energy performance) are selected for
further development. As the building concept reaches the detailed design stage, those simple
sketch lines are being replaced by specified dimensions, construction materials, mechanical
systems and schedules etc (i.e. detailed simulation inputs) to paint a clearer/fuller picture of the
proposed design.

The relationship between the design lessons of performance evaluation tools applied at sketch
and detailed design phases has to be consistent in order for these tools to be viewed as
providing plausible design advice/feedback. Hensen (1994) says:
“Building energy simulation is considered to be a potentially powerful tool for
decision support in energy efficient building design.”

All these building examples of sketch to completion have a common design characteristic: they
all have heavily glazed façades. It is apparent that glass is a popular method of construction. In
this research, COMFEN is used to evaluate sketch ideas of highly glazed commercial building
design; it is an early design tool that focuses on sketches of fenestration design (i.e. help
designers identify the optimum fenestration design for a building).

13
1.4 Research Overview

This first chapter is an introduction to the performance sketch idea and the research overall.

The next chapter looks into the literature of early design tools and their users. It discusses the
evidence for the influence of early design tools. Also, this chapter focuses on commercial
building façade design (aesthetics & energy performance) because of its role in occupants’ well-
being and the overall building energy consumption.

The third chapter describes the methodology of this research, from calculation
methods/techniques used in simulations to structured interviews with practitioners. It describes
the three-method approach, and how each method complements the others. The methods are:
1. Introduction of a Performance Sketch Tool to Practitioners
2. Consistency Test of a Performance Sketch Tool
3. Evaluating the Adequacy of the Sketch within a Performance Sketch Tool

The fourth chapter discusses the practitioner feedback from the use-case studies and the
results derived from COMFEN and EnergyPlus simulations.

The fifth chapter contains a list of performance sketch tool specifications based on the use-case
studies of chapter four.

The Conclusions chapter justifies the overall outcome of this research and examines each of
the research questions in relation to the analyses of chapter four and the aims of the research.

The Future work chapter describes the potentials of Performance Sketch tools like COMFEN
and reports research questions that require further testing and investigation.

The Appendices chapter includes all COMFEN (single-zone) simulation outputs, EnergyPlus
results, façade material properties, and original interview feedback from practitioners.

14
2 LITERATURE REVIEW

This chapter looks into the literature of building energy simulation tools in relation to their role
during early design and detailed design stages of building construction. There is a need for
energy simulation tools because architecture practitioners utilise them to estimate energy
performance of building designs. The chapter identifies the audience (users) of energy
simulation tools and determines their simulation needs during different stages of design.

Also, the subject of heavily glazed façades is addressed. Section 1.3 illustrated a range of
building designs that consist of highly glazed façades; designs driven by the conceptual sketch.
The literature discussing the energy impact and the reasons why this method of construction is
so popular in the commercial sector is extensive. The use of energy simulation tools is
significant in this discussion because a building design is (ultimately) tested by these tools
during design. This is where the aesthetic decisions about the façade which determine the
overall energy impact can be evaluated.

2.1 Users of the Early Design Tool

Generally, the audience for early design tools for building energy performance are architects
and engineers. The accuracy of these tools is essential in order to deliver a successful building
project. Clevenger & Haymaker (2006) state that energy modelling accuracy (comparison
between two models in the same simulation software) ranges from +/- 10% ~ 40% for non-
residential models and general industry consensus is that comparisons of predicted
performances are more useful than the absolute values themselves.

Many developers have the intention of knowing how a building design performs before
construction begins, which is why practitioners (widely) use early design tools during the early
stages of design; these tools allow practitioners to identify and evaluate the best concepts to
show the developer which ideas are worth further development. Bordass & Leaman (1997)
state:
“Clients, naturally enough, hope that a new building will magic many of their
building related problems away and leave them to concentrate on their businesses.”

The intention of computer energy simulation is to create a virtual representation of the proposed
design and determine its energy performance and environmental impact. However, various
assumptions have to be made by the user (practitioner) in terms of building materials, lighting
loads, building schedules and HVAC system which all require input data into the software.
These design variables influence the results depending on the simulation assumptions made
and the user’s familiarity with the simulation tool. Therefore users of design tools need to have
great understanding of just what is being simulated and the software’s intended purpose. The
user must recognise what the design tool is able to do and cannot do; using the appropriate tool
to assess the proposed design is vital because a tool that is not intended for the desired type of
analysis could produce estimations that are misleading to the user.

15
Like an architect’s sketch, early design tools evaluate designs quickly. Detailed design tools
require a much larger amount of data entry; they are therefore more time consuming.

Anytime during the early design phase the architect’s concept sketch can be returned to the
drawing table. Design alterations have to be made until the client is satisfied. The ‘Performance
Sketch’ idea needs to be relevant to this early design stage; the capability of examining multiple
designs quickly and accurately has to be time efficient for practitioners.

It is important to note that a simulation tool should work in favour of the user and not the other
way around. The users of early design tools often base their opinion on software usability and
capability (i.e. user-friendliness & options of analysis). The following sources express users’
opinions of building energy simulation tools:

Attia, Beltrán, et al. (2009) mention:

“Architects are looking for tools that can support sustainability design decisions and
make detailed comparisons between different building design and equipment
measures.”

Tianzhen, Jinqian, & Yi, (1997) state:

“From the perspective of many architects, most building performance simulation
tools are judged as too complex and cumbersome.”

Attia, Beltrán, et al. (2009) say:

“Most architects who use building performance simulation tools in design practice
are much more concerned with the (1) Usability and Information Management of
interface and (2) the Integration of Intelligent design Knowledge-Base.”

And, Punjabi & Miranda (2005) say:

“However, despite the proliferation of many building simulation/energy analysis
tools in the last ten years, architects and designers are still finding it difficult to use
even basic tools.”

“User-friendliness” is reported generally as playing a huge role in determining how the user
perceives the design tool. It is vital that the design tool is user-friendly because then the user
can integrate their professional knowledge and experiences into simulations of the proposed
building design. “In order to create a simple building energy simulation tool that could help
architects and engineers to calculate the impact of design choices the program had to be: highly
user-friendly; need a minimum of data; and be very fast.” (Gratia & De Herde 2002).

To Shaviv, Yezioro, et al. (1996) user-friendliness is ultimate flexibility in data entry:

“Most parameters are not assigned specific values early in the design process.
Therefore, the designer is free to choose almost any value for the parameter of his
current concern. As the design progresses, the previously assigned values may
restrict the designer's freedom in choosing the values that are assignable to
parameters considered later.”

16
User-friendliness is defined by this statement as ultimate flexibility. However, others argue that
more flexibility means the tool becomes more complex; ultimately, it could be harder to
understand and use. Bambardekar & Poerschke (2009) state that sophisticated simulation tools
are used to test both macro and micro level performance parameters; but they have the
following disadvantages:
• Steep learning curve and time intensive application
• Requires large amount of input data

Practitioners have virtually no limitations in terms of expressing their architecture or energy

concepts during the early design stage because these are often firmed up during the detailed
stage of design. The practitioner wishes to use early design tools to assess various design
options with only few restrictions. Early design tools also require minimal information for the
user to perform design simulations because during the early design stage the specified design
information is not yet available (e.g. HVAC system, lighting load, & building schedules etc).

Each simulation tool has modelling restrictions/limitations. This research proposes there are two
types of early design tool:
• Traditional early design tool: limited data entry is required due to restrictions on
the design situations to which the tool can be applied: for example, limited
building size, type, location, system.
• Sophisticated early design tool: allows access to a detailed design simulation
tool is made available during the early design stage via a simplified model (i.e. a
performance sketch tool).

One example of a traditional early design tool is a set of rules of thumb for building design
based upon an experienced designers’ documentation of their previous designs; or upon
summaries of the results of multiple uses of a simulation package on multiple variants of a
standard building. Ultimately, the predictions or rules of thumb derived are dependent on the
nature of the standard building.

A sophisticated early design tool provides the user with the modelling flexibility of a
comprehensive simulation package but reduces the complexity of data input. The user has the
ability to assess a simple model with the option to adjust the essential ‘detailed’ simulation
inputs (e.g. occupancy, lighting & interior equipment loads) during the very early stages of
design.

The sophisticated early design tool allows the user to set the desired simulation inputs rather
than working with defaults or standard designs not relevant to their site or client. Also, modelling
flexibility allows the designer to explore a wider variety of concepts than the traditional tool.

It is essential to incorporate past experience (lessons learnt) to a new building project. In this
sense every professional designer (Lain, Hensen, & Zmrhal 2009) performs some kind of
optimization in each stage of the design process in order to ensure a good indoor environment
quality while limiting or minimizing energy consumption. Making an actual design decision relies

17
on the designer’s ability to explicitly represent, and then reflect upon, the desirability of the
performance of a certain constellation of form, function and context (Petersen & Svendsen
2010).
“It is a matter of applying our technical knowledge and our long years of experience
logically to create concepts for new and refurbished construction that will lead to
low-energy buildings.” Hermann Kaufmann (Hausladen, de Saldanha, & Liedl
2006).

These authors suggest that to design a sustainable building, all environmental factors have to
be examined (taken into consideration) during the early design stage. They suggest use of early
design tools is the most practical and effective method of testing for getting an indication of the
proposed design’s overall energy performance.

Some tools employ ‘simplified’ methods that address specific perceived needs of the early
design phase while others adopt complex first-principle based engineering algorithms that can
meet detailed design requirements (Lam, Huang, & Zhai 2004).

“Another important task is the issue of quality control, both for input and output. Input issues
regarding internal loads are the most important since they are based on expertise (i.e. the data
is not available from the product specifications). The output has to match in terms of expected
results and common benchmarks, which are also based on experience.” (Brahme, O’Neill, et al.
2009). Choosing the appropriate design tools for the desired analyses is very important:
“A frequently encountered problem by engineers who would like to perform a
simulation is that there is no single simulation environment that can cover the whole
range of problems at hand. Certain performance aspects are available in one
package while other aspects are only available in another package.” (Hensen,
Djunaedy, et al. 2004).

Some design tools require the user to have previous simulation experiences simply just to
understand its basic principles. Not to mention the output data, the ability to understand and
incorporate the predicted information to the design process depends on the user’s level of
simulation knowledge and their familiarity with the software.

Once the user understands the software principle it will become clearer to them just exactly
what the predicted design message is indicating (i.e. simulation outputs) and really experiment
with a variety of design options with confidence and trust in the software. Commission of the
European Communities (1993) states:
“The practical efficiency of building energy simulation tools is dependent not only on
the facilities offered by the tools and the rigour of the underlying (dynamic thermal)
calculations but depends also on the skills of the user in terms of abstracting the
essence of a problem into a model, choosing appropriate boundary conditions,
setting up simulations and interpreting the results.”

18
Tools with built-in simulation examples allow the user to learn and understand the functionality
side of the software (i.e. capabilities). As the user becomes more familiar with the software then
the architectural drawings of the proposed design can be transferred into the virtual
environment. It is essential to select the right tool for early design analysis. Having a design tool
that is capable of exploring crucial design variables from an early design perspective is
important to architects and engineers.

Early design tools have various audiences. Apart from the general users (e.g. architects &
engineers), there is one other vital audience group and that is the clients. The information
extracted from these early design tools must be presented to the client with clarity, so they
understand the performance of the proposed design. Early design tools should therefore have
the capability of providing the design information that practitioners require for client
presentations. Often the output data predicted by the simulation software needs to be translated
into graphs and tables and other information that the non-expert can understand. The process
of creating those graphs and tables could be time consuming (and tedious) therefore tools that
automate the provision of useful graphical information to represent the simulation data is
extremely relevant for client meetings and progress reports.

Various authors suggest the importance of graphical output from an early design tool. “To
understand the architects’ perceptions about existing tools and the importance of using them
during design phases.” (Attia, Beltrán, et al. 2009). “It would be desirable to develop
visualizations that would better facilitate a qualitative understanding of the design performance
to the user and provide appropriate guidance in the context of early design decision making.”
(Lam, Huang, & Zhai 2004).
“Architects need a tool that provides graphical representation of simulation input
and output, simple navigation, flexible and customizable control, in addition to
intelligent default features.” (Attia, Beltrán, et al. 2009).

The relationship between software developers and users of their software is an area that lacks
research and understanding. There needs to be a stronger link between the theory side of
thermal calculation and real world (professional) practice. Understanding what
practitioners/simulators need is very important for the role of a software developer. What may
seem plausible to the developer at times might not be plausible to the user. It is crucial to know
exactly what the users need during the development phase of these tools. In relation to building
design, this research had access to information regarding to client desires and early design
simulation needs through practitioners’ interview feedback (i.e. use-cases).

19
2.2 Relevance of the Early Design Tool

The performance sketch tool uses a detailed design tool in a ‘sketch manner’ to evaluate a
building design (i.e. like an architect’s sketch pad). Simplification of a detailed simulation
application allows the user to sketch building ideas quickly and assess the essential design
features during the very early stages of design (e.g. fenestration design & mechanical loads).

When writing about early design tools, it has been assumed by many of these authors that a
complex computer simulation tool requires a complex and complete definition of a building; the
simulation process is (often) slow and tedious due to the complexity of the required inputs.
“The complexity of the data input is reduced by providing smart default values for solar systems
and building elements, to broaden the use of this tool in the architects’ community.” (Witzig,
Foradini, et al. 2009). They suggest that the use of smart default settings could replace the
process of manual (simulation) data input entry which generally speaking is time consuming and
complicated. This complexity has been perceived as a barrier to the use of complex computer
simulation during the very early stages of design.

Ochoa & Capeluto (2008) state that early design decisions are based on vague ‘‘ideas’’ that
cannot be evaluated with tools that rely on exact data. Many design tools have built-in example
files within its application. Default settings used within the example files are very useful to
designers because it allows them to perform simulations of their design concepts with very
minimal information required. The advantage of smart default settings is that the user can alter
these values to their liking. More detail can be incorporated into the simulation model as the
design progresses. In other words, the user can adjust the settings to the desired values
according to the finalised building specifications.

It is important to note that the performance sketch tool is of a different nature to the traditional
type of early design tool proposed by Ochoa & Capeluto. The basic philosophy of the
performance sketch tool is to use a comprehensive design simulation package in a much
simpler manner. This provides the user with the power of a detailed design tool during the early
stages of design, which comes with more modelling flexibility than the traditional approach
where predictions derived are dependent on default settings within the application. The
sophisticated early design tool allows the designer to explore various concepts in a sketch
manner. The advantage of making some complex applications available from day one is that the
sketch models can be duplicated, re-used and altered within the detailed design tool as the
design progresses into the development phase.

Goulding (1993) noted that it is widely acknowledged that the best opportunity for improving a
building’s energy performance occurs early in the design process when basic decisions are
made. Many practitioners in today’s building industry use some kind of early design tool to
estimate the performance of their proposed design. The predicted data is used to assist the
design development process. Therefore, the accuracy and relevance of these design tools have
a significant influence on the user’s trust.

20
Using computers to carry out load calculation and energy simulation is an important part of
building design and energy efficiency research (Hui 1998). Once a project reaches the detailed
design stage alterations to the building design may be extremely difficult. Therefore
performance predictions by an early design tool are significant. If the basic requirement is not
fulfilled (Eisele & Kloft 2003) all other attempts at creating a contented atmosphere through well-
designed space are doomed to fail.

“Decisions made in the very first stages of a building’s design often have a
significant impact on energy efficiency and internal environment of the building.
Although many buildings have energy efficiency strategies embedded in their
conceptual design, it is seldom that these concepts would be fully analysed at the
initial design stages.” (Pollock, Roderick, et al. 2009).

Design decisions are influenced by early design tools. Early design tools not only assist
practitioners designing sustainable buildings but also allow designers to understand and explore
the costs and benefits of alternative concepts. During the early stages of building design a
number of decisions are made which have a strong influence on the performance of the building
throughout the rest of the process (Petersen & Svendsen 2010).
“Moreover, the penalty for not addressing climatic responsive design issues early in
the process is that opportunity will be lost to make significant savings by relatively
simple adjustments to the design. Increasingly sophisticated or costly efforts are
needed to save energy.” (Goulding 1993).

Lomanowski & Wright (2007) state that few tools exist that can aid the building designer in
quantifying the impact of window shading on building loads. They suggest there is a clear need
for an explicit treatment of window shading layers in building energy simulation especially during
the early design stage. Many have proposed the essence of getting the building design right
from day one; it should be common practice and take advantage of determining the ideal design
before it is too late (i.e. during the stage when design is flexible not fixed). They emphasise that
there is a desire for software capability and usability improvement, from a fenestration design
analysis point of view.

When using early design tools to examine the energy performance of a proposed design, it is
essential to differentiate design concepts that work and ones that perform inadequately. The
predictions from these tools assist people with vital decisions that need to be made as the
project progresses; predictions which help identify the best solutions to meet the project
requirements.
“An early analysis will help to establish the merits each design solution offer and
help designers develop an optimum solution that meets all required criteria. Where
problems with occupancy comfort can be identified early in the design process,
passive or low energy solutions can be developed, tested and implemented into the
final design.” (Pollock, Roderick, et al. 2009).

21
Attia, Beltrán, et al. (2009) state the importance of realizing the impact on energy performance
and cost from the early stages of design.
“The value of computer simulation when used in early conceptual design lays in the
opportunity to quantify the effect of design decisions on energy and comfort thereby
reducing the time needed to turn-around design iterations.” (Lain, Hensen, &
Zmrhal 2009).

The general goal of energy analysis is to model the proposed design and see if it meets the
desired energy performance and comfort levels (standards/codes and client requirements).
Comparing the energy performance of various designs allows practitioners to narrow down to
the best design options. Client dependent, the best concept(s) is later carried into the
development phase of the project for further testing.
“Simulation is much more effective when used for comparing the predicted
performance of design alternatives, rather than when used to predict the
performance of a single design solution in absolute sense.” (Hensen, Djunaedy, et
al. 2004).

There are many advantages using an early design tool – a virtual representation - because it:
• provides an early prediction of the proposed design’s energy performance (e.g. EUI)
• can identify design problems earlier on in the project, so it can be solved/remedied
• helps the client and practitioners with the decision-making process

Architects rely on early design tools, they use early design tools to assess their designs and
help them with decision-making. However, when it comes to building energy performance,
predictions by an early design tool have to be plausible and consistent. The need for early
design tools to be reliable is essential because it is too late to make changes after the building
is completed; where changes to remedy defects could cost thousands of dollars.

“Practitioners need early stage, strategic design decision support tools. In the area
of indoor environment, building physics and building systems complex interactions
exist which are very difficult – if not impossible – to capture and represent in rules
or other forms of explicit knowledge for use in knowledge based systems. This is
the main reason why many current knowledge based tools are often restricted to
single issues.” (Hensen, Djunaedy, et al. 2004).

The authors suggest the importance of selecting a relevant tool for the intended design analysis
because each software application has its own simulation purpose(s).

22
Often when exploring a new building design that has not been simulated in the past, estimates
and ‘educated assumptions’ have to be made based on the designer’s past experiences.
Brahme, O’Neill, et al. (2009) state:
“Although many of the strategies can be simulated with current tools, there are
some that have to be approximated, some have to be drastically simplified both for
ease of modelling and time reasons.”

The authors raise a key issue with some simulation tools being “drastically simplified” for user-
friendliness and time efficiency purposes. The tendency with overly simplified tools is that they
are not flexible, which restricts the designer from exploring various ideas from day one. This
contrasts with the needs of a designer at the beginning of a building project when they should
be identifying and evaluating the best solutions to carry into the detailed design stage.

Design parameters usually come from the project brief or the building budget. However, the
simulation tool should never shape how the building is designed due to its software limitations.
The designer needs a tool that is capable of testing the client’s desired concept(s). The key is
‘software capability’ because it is important to distinguish what the tool can simulate and what it
cannot. Lam, Huang, & Zhai (2004) suggest:
• “The user interface should be designed such that it is familiar, cognitive and
compliments the concepts and processes of architectural design and energy
modelling.”
• “For a tool to be beneficial and remain relevant throughout the building delivery
process, it would be advantageous if it is developed based on comprehensive and
fundamental principles in modelling the building-environment interactions.”

“It is essential that the simulation tools include an interface that supports such a knowledge-
base. A knowledgebase that contains descriptive explanations, examples and procedural
methods for determining appropriate installation and systems, e.g. guidelines, case studies,
strategies etc.” (Attia, Beltrán, et al. 2009). The authors state the importance of simulation
interface. In this sense a simulation interface with clear documentation can help the user to
understand software application and its outputs. It is essential for the user to understand the
principles of a design tool; this can be achieved through the simulation interface proposed by
Attia, Beltrán, et al.

“Computer modelling and simulation is a powerful technology for addressing interacting

architectural, mechanical, and civil engineering issues in buildings.” (Hensen, Djunaedy, et al.
2004). There are many current design tools that simulate various environmental factors in
relation to building design (e.g. over-heating, glare & heat loss). However, it is essential for a
design tool to be capable of simulating various design options that the user desires to assess; to
address design aspects that influence the building’s overall performance.

23
Building simulation tools allow today’s practitioners to create a virtual model of their building
design and assess its performance.
“A design tool would go beyond calculating the energy performance to address
other building design considerations like comfort, economics, and aesthetics. Also,
a design tool would help its users formulate appropriate design criteria and improve
building performance as the design evolves.” (Papamichael & Ellington 1994).

It is clear that with the assistance of early design tools the designer will be able to identify what
works well and areas that require improvement within a proposed building design. It is
extremely important that practitioners understand the energy performance of their designs, and
deal with issues/defects earlier on when mistakes can be fixed without major confrontation.
Therefore these tools are the most practical solution for answering those early design questions
by architects and engineers.

The performance sketch analysis approach of LBNL’s COMFEN is of a similar nature to the
user-friendly simulation tool (OPTI) proposed by Gratia & De Herde (2002); it is fast and
requires minimal data. COMFEN focuses on the fenestration design of the building with
simplified applications of a detailed simulation engine.

“With increasingly affordable computing power, it is argued that energy modelling tools should
adopt rigorous physics and engineering-based algorithmic principles in the computational
prediction of energy performance to ensure acceptable results.” (Lam, Huang, & Zhai 2004).
The authors suggest that from an early design assessment perspective there is a need for
ensuring the plausibility of energy predictions derived from early designs tools.

The performance sketch tool allows the user to explore new ideas during the building design
development phase, and has the added advantage of being compatible with a detailed design
tool. Quick sketch analyses can be done parallel to the detailed comprehensive model if
alterations are required during the detailed design stage; the sketch approach is time efficient.

24
2.3 The Relationship between Early Design Lessons and Detailed
Design Lessons

In this research COMFEN is the selected example of an early design (Performance Sketch) tool
that determines the energy performance of various façade sketches. These early design
messages are incorporated into detailed simulations using EnergyPlus.

Architects and engineers depend on energy simulation tools to provide them with predictions on
how their proposed design performs before it is built. Hui & Cheung (1998) wrote:
“The purpose of energy calculation is to estimate the annual energy consumption of
buildings so as to provide information for energy and economic analysis which aims
at improving the building design.”

As a practitioner you would expect the best proposed concept developed during the early
design stage to perform in a similar/plausible manner as the comprehensive model when
specified details are incorporated into the simulations. Early design predictions have to be
consistent with results derived from the detailed design stage. If the early design message is
inconsistent there is a possibility of misleading practitioners during the early design stage;
defects that occur during the detailed design stage could be expensive to remedy. Therefore,
the relationship between early design tool and detailed design tool is significant.

The transition process from the early design stage of a building to the detailed design stage is a
significant phase in a building project. Ultimately, the decisions made during the early design
stage determine/shape the final performance of a building project.
“The early stages of this process characterize themselves by a constant search for
a design direction. But as demonstrated by specialists in design methods, decisions
taken in those moments can determine the success or failure of the end product.”
(Hari 2001).

Early design tools are suitable for the preliminary (early) design stage because simulation
settings are often based on defaults until the building design specifications are determined,
which are specified later in the project (i.e. development phase). The ‘sketch’ design
considerations during the preliminary design phase is important because of its effect on the
detailed design stage.
“The fragmented nature of the building process, in which no member of the design
team considers the overall optimization of the indoor environment, further
compounds the problem. Since the façade and fenestration design relates to
different aspects of building performance (heating, cooling, lighting) and human
comfort (thermal, visual), an integrated approach should be followed from the early
design stage.” (Tzempelikos, Athienitis, & Karava 2007).

25
Fabrizio, Corrado, & Filippi (2010) state that it is well known that the potential benefits of the
design inputs taken at the design concept stage are much higher than the benefits of design
choices taken at the design development and construction document phase. “The cost of
implementing concepts to improve the energy performance of a building is also lower at the
earliest stages.” (Lewis 2004).

Early decisions made in a project play a huge role in the relationship between early and detailed
design. All aspects of the design must be examined as early as possible, from the (exterior)
building envelope to comfort of occupants in the internal environment. The ideal is to carry the
best concept(s) from early design analysis into the detailed (development) stage and apply
specified details into the simulation model for in-depth analyses before finalising the building
design.

Some simulation tools can be difficult to understand because of software complexity. Hong &
Tianzhen (1997) state that the procedure of building energy simulation is not simple and there
are a lot of factors to consider. Architects and engineers may not able to master it without
proper training beforehand. In general, the procedure of the simulation will be as follows:
a) mastering of the simulation tools;
b) description of the building design and the assumptions;
c) preparing of the simulation inputs;
d) carrying out of the simulation; and
e) interpretation of the simulation results.

Walton (1989) & Feustel (1989) pointed out most existing tools are either too “simplistic” or too
“complicated” to provide effective design support. The simplistic ones are often too limiting, in
that they apply only to highly generic situations regarding building geometry and operation.
Twenty years after Walton & Feustel’s statement the situation has not changed; the following
are others suggesting that simulation tools are too complex for early design simulations:
• “The currently available design advice tools tend to focus on the development of a
platform for the evaluation of alternative designs rather than giving actual design
advice.” (Petersen & Svendsen 2010).
• “Most tools are dedicated to evaluate and model a certain finished alternative, not
to suggest and evaluate different design options and directions. This implies fitting
an idea to the modelling tool, thus filtering out information that could be useful or
distorting the process.” (Ochoa & Capeluto 2008).
• “The sophisticated ones typically demand too much information, time, and
expertise to be helpful to the primary building designers (usually architects) in the
early stages of building design.” (Gan 2000 and Papakonstantinou, Kiranoudis, &
Markatos 2000).

From an early design perspective, to sketch what needs to be accomplished is more important
than the specific details. The detailed specifications can be applied once the overall concept
has been established.

26
Obviously, the simpler the tool, the easier it would be for the user to perform design simulations.
When less detail is applied, the predictions are simply a quick indication of the design
performance. This is ideal for early design analyses because practitioners can simulate various
designs quickly and narrow down the proposed options for development phase.

However, if more specified details are incorporated into the simulations, the more relevant these
predictions would be to the completed building; the only issue is that those kinds of information
are not fixed (determined) during early design. Lam, Huang, & Zhai (2004) say:
“The conceptual approaches adopted and technical implementation of these tools
varies significantly. Some tools employ ‘simplified’ methods that address specific
perceived needs of the early design phase while others adopt complex first-
principle based engineering algorithms that can meet detailed design
requirements.”

Also, Petersen & Svendsen (2010) point out:

“Making informed design decisions requires the management of a large amount of
information on the detailed properties of design options and the simulation of their
performance.”

The key issue studied in this research is the need to develop a trust in the relationship between
the predictions of tools used during the early and detailed design stages. Possible design
defects investigated and solved during the conceptual stage must remain design issues for
which the solutions are valid at the detailed design stage. “The tendency to adopt abstraction
and rule-of thumb approaches in an attempt to meet the time and resource constraints
encountered in early design should be avoided.” (Lam, Huang, & Zhai 2004).

Early design predictions produce design concepts which are incorporated into detailed
simulations for further testing with specified building specifications (e.g. HVAC & lighting).
Brahme, O’Neill, et al. (2009) wrote:
“Simulation tools are mostly used during the detailed design stage when most of the
decisions regarding building massing and system types are already made. The
tools in this case allow one to understand the impact of various building and system
component efficiencies.”

The urgency to meet the initial design deadline during the ‘sketch’ design stage can often rush
the decision-making process. It is paramount to get the design correct from day one and avoid
shortcuts that could potentially lead up to further problems somewhere down the line during the
design development phase; by that stage additional cost to remedy those design defects is
inevitable.

27
The following diagram proposed by Kalay (1999) describes decision-making during the early
design stage before the project progresses onto the detailed design phase:

Figure 15 – Workflow and subtasks in performance-based design (Kalay, 1999)

Figure 15 portrays the notion of getting the preliminary design right before moving onto the
detailed design stage. No further project development is undertaken before the desired design
performance has been met. To meet the desired performance is the responsibility of the
practitioners who use early design tools to predict performance of building designs, which also
need to meet the design concept that the client is looking for. When the project enters the
detailed design stage, any mistakes discovered can be extremely difficult and expensive to
remedy; not to mention time consuming. Petersen & Svendsen (2010) state:
“The workflow in Figure 15 is an attempt to improve the ability of the designer to
facilitate the design activities in the conceptual design phase.”

28
2.4 Commercial Building Façade Design

The discussion in Section 1.3 shows a range of building designs typical of modern high rise
buildings where highly glazed façades are popular. This research has focused on fenestration
design for these types of buildings because they are clearly popular and because their
performance is widely debated. More than many other buildings, this performance in use is
critically dependent on design decisions made during the early design stages.

“Fenestration systems are major elements in the energy balance of office buildings,
and their design is a critical part of the building design process. The selection of
glazing materials, systems and window treatments such as overhangs and shading
devices can have major impacts on building energy use.” (Persily 1993).

It is possible to find various references to the desirability of glass as an envelope material in

commercial building façade design, even though this type of construction method can require an
immense amount energy to maintain comfort indoors. The system is perceived to enhance the
building façade’s appearance and provide occupants with natural light in the work environment.
For example:
• “Glass is one of the best construction materials for giving a building a
homogenous appearance. This look is associated with dynamism, precision,
prestige and progress. Also: The trend for as much transparency as possible in
architecture embraces the wish to dissolve the outer skin and make the building
appear light and open.” (Hausladen, de Saldanha, & Liedl 2006).
• Lomanowski & Wright (2007) discuss the importance of solar gain management in
energy efficient building design: “Given the current architectural trend toward
highly glazed façades in commercial buildings.”

Energy simulation tools can determine the most energy efficient concepts. Therefore, if the
client decides to have a ‘glass box’ (building) design these tools can assist the design team in
establishing the best solution. Early predictions of heat loss/gain derived from energy
performance modelling of the building design can help the designers to determine the best
combination of shading system and type of glazing that would minimise the energy consumption
and still maintain the desired exterior appearance. The practitioners’ responsibility is to integrate
any client-desired features into the building design; if you have to build this way (e.g. for
aesthetics purposes) it is essential to identify energy and cost efficient concepts from the early
stages of design.

29
Heavily glazed façades are neither bad nor good. They are just one of many building features
that have to be simulated and evaluated during the design phase. It is important to note, in a
simulation approach there should be very few shortcuts or assumptions connecting general
building design features and performance. The early decisions made are very influential to the
project overall.
“Often the building services engineer will be involved later in the project where
many of the decisions over fabric, shape, layout, glazing and orientation have been
made and fixed. Therefore by this stage the ability to utilise the most appropriate
passive measures may have been heavily restricted and mechanical conditioning
systems are needed to maintain occupancy comfort.” (Pollock, Roderick, et al.
2009).

Environmental Factors which are generally agreed to have a large impact on a commercial
building’s energy consumption and occupants’ wellbeing and which are governed by the amount
of glass in a building façade are:
• Building over-heating potential
• Heat loss of interior space
• Glare from daylight/sunlight
• Glare from neighbouring heavily glazed buildings

In 2006, the US Department of Energy reported that buildings consumed 40% of the energy in
the United States and 12% of the water (DoE 2007). And, Prasad & Bhat (2005) say:
“Approximately one third of our primary energy supply is consumed in buildings.
Consequently, buildings are a primary contributor to global warming and ozone
depletion. Therefore, achieving better energy efficiency in buildings has become
one of the world's major challenges.”

These authors suggest the negative effects of designing building façades that are dominated by
glass for aesthetics purposes. Whereas the sustainability side of the design can sometimes be
overshadowed by the overall building image. Bannister (2009) emphasised the significance of
building design during the preliminary design phase:
“…there are some buildings that are doomed from day one to perform poorly. In
some cases this is because the basic design is poor, but the problem can also
affect buildings that have apparently reasonable design and construction, at least at
the macro level.”

30
If the tendency is to clad the building envelope with glass to enhance its aesthetic appearance,
then it is vital to examine all environmental factors during the early design stage. Glass
buildings provide occupants with daylight and connection to the exterior environment. In this
sense glass façades are perceived as an important design feature in a building:
“The importance of glass as a cladding material cannot be underestimated. With the
exception of plastics, it is the only transparent cladding material, providing
daylighting and connection with the outside. It is rare client who desires less light-
transmitting glass on a project.” (Brock 1948).

“Numerous office buildings of the 1980s were designed to isolate the internal conditions from
the outdoor climate as far as possible; this at the cost of high energy consumption.” (Voss,
Herkel, et al. 2007). The authors back up Brock’s opinion on the importance of glass cladding. It
is suggested that for this reason it is important to retrofit old buildings to achieve sufficient
access to daylight that can help minimise electric lighting energy consumption (i.e. optimum
Window-to-Wall Ratio).

There are many aesthetic and other reasons encouraging architects to build with glass façades.
Despite the environmental factors and energy impact, there are several reasons why glass
façades are so popular (by practitioners’ demand) in the commercial building sector:
1. Aesthetic of the building is appealing
2. Access to natural light provides occupants with the indoor-outdoor connection,
which result in higher productivity (healthy work environment)
3. Speed of construction allows the tenant to move in the property as soon as
possible
(e.g. unitised system – curtain wall construction)
4. It enables the firm to secure building projects

These reasons have a significant influence on the decisions made by the design team and
client, hence the familiar trend of glass buildings in cities around the world. Architects are
almost forced to design this way to secure projects:
“Building façade design is developed at a very early stage of the design process. It
is often the basis for the award of the contract to a particular firm of architects or
developers.” (Kolokotroni, Robinson-Gayle, et al. 2004).

31
Figures 16~18 show examples of highly glazed commercial buildings. What they have in
common is that their ‘glass skin’ building aesthetics make them a landmark within their own city.
There are two main aspects that make these buildings stand out, their scale and geometry.

Scale: towering expression of Taipei 101 (height of 508m) and International

Finance Centre (height of 416m)

Geometry: angled façades of the Audi Forum building and Taipei 101’s
bamboo exterior appearance

Ultimately, these two design aspects express the architectural idea envisioned by the design
team.

Figure 16 – Exterior view of Taipei 101 in Taipei, Taiwan

Figure 17 – Exterior view of the Audi Forum building in Tokyo, Japan
Photographs © Chi-Yao Hsu

32
Figure 18 – International Finance Centre in Hong Kong. Image © (Crosbie, 2005)

Despite the poor thermal properties of glass (which cause overheating & heat loss), these
examples (Figures 16~18) are demonstrations of the popularity of glass amongst clients and
design practitioners. Often, designs of heavily glazed (high-rise) commercial buildings rely on
mechanical systems to condition the interior spaces. In other words, as long as the building
aesthetics meet the design brief in relation to image and prestige other design aspects (such as
energy use and envelope thermal performance) can be compromised. Many people argue that
a simple glass façade causes internal comfort and energy performance issues:
• “Although the user enjoys more of a view out, there is an increased risk of glare
and overheating in summer.” (Hausladen, de Saldanha, & Liedl 2006).
• “In terms of energy consumption, much of the existing commercial building stock
is made up of multi-storey, highly-glazed, thermally-lightweight developments
that are totally dependent on non-renewable energy for heating, cooling and
lighting.” (Fernandez 2008).
• “There is just too much glass used on these curtain wall systems, all because of
the aesthetic and this makes it very difficult to enclose conditioned air; the
building relies on the mechanical systems to maintain comfort. Fenestration and
shading systems have a major impact on visual and thermal comfort in
perimeter spaces but also on energy consumption, peak loads, and possibly
HVAC system sizing.” (Tzempelikos & Athienitis 2005).

The authors suggest that heavily glazed buildings (like Figures 16~18) rely on mechanical
systems to condition its internal environment to keep the building occupied from day to day.
33
This is not a new issue and the debate about it is ever growing. Hausladen, de Saldanha, &
Liedl (2006) wrote:
“The objective in the design of façades is to find the optimum compromise between
the specific location and the various requirements of the planned building use.”

The reality is that research of this type might not have the ability to control how people design
their buildings, but it can forward the message to designers that the use of simulation tools from
day one will emphasise just how much energy glass-greedy buildings can (potentially) chew up
each year and throughout their lifetime. When designing a building it is argued that the analysis
should be focused on long term use, not short term. Often, long term effects are ignored, as
Flanagan & Norman (1983) stated:
“The difficulty of forecasting future events in the life of a building and its
components, which is part of the reason why LCC considerations are often
overshadowed during the design process.”

“Transparent façade components allow for passive heating of the interior through insolation. Far
more frequently, however, solar heat gains will contribute to overheating and increased cooling
loads. The total window area on the exterior wall and the efficiency of the shading system are
key factors.” (Eisele & Kloft 2003). There are ways (potentially) of making all glass façades work
at least as well as other systems if they incorporate active shading device etc. The following are
some solutions proposed by Hausladen, de Saldanha, & Liedl:
• “The use of natural light has a considerable influence on the energy demand of a
building. The direct saving comes from the reduction in electricity used for
lighting.”
• “External solar screening is the most effective system because the solar radiation
is blocked before it can reach the façade. However, initial and maintenance costs
are higher because the system is exposed to weather and wind.”
• “Internal solar screening is protected from the weather, can be operated in all
wind conditions and provides glare protection.”

Winfried Heusler (Hausladen, de Saldanha, & Liedl, 2006) adds that for glass façades to work
they require some type of active shading system to adapt according to the weather:
“Building skins that have been properly designed to meet their climatic and usage
requirements react to changing outside conditions as a semi-permeable membrane
with dynamic properties instead of presenting a rigid, impenetrable barrier between
the room and the outside environment.”

Also, there are cost implications of active glass façades. However, Hausladen, de Saldanha, &
Liedl (2006) point out:
“Room climate can be considerably improved at little extra cost if the characteristics
of a façade arising from its orientation are taken into account in the strategies for
providing solar screening and adequate ventilation.”

34
Hausladen, de Saldanha, & Liedl (2006) argue building façades should adapt to their
environment so the complexity of mechanical systems can be reduced:
• “Façades must be able to react and vary themselves in response to the dynamic
outdoor climate and ever-changing indoor climatic conditions.”
• And, “The more adaptable the façade is, the less complex the building technical
systems need to be and the lower the associated energy demand is.”

However, Leaman & Bordass (1999) argue that active façades which are well-tuned to their
climate could have potential failure risks:
“Bigger and more complex buildings demand subtler strategies for managing this
complexity and different design strategies and technologies to support them. Where
this is successful, performance gains are possible, but where management does
not properly compensate for the extra diligence that technology needs, chronic
problems usually result.”

And: “No begged questions, Keep it as simple as possible, but not more so; Make it
adaptable; If in doubt, leave it out; and What if...so what?; can be appropriate
rallying cries.” (Bordass & Leaman 1997).

One way to avoid this is to get better information during the early design. This can be achieved
with an early design tool that produces advice/feedback consistent with that produced later in
the design process from a detailed design tool. Bordass & Leaman (1997) state that designers
sometimes collude in this fantasy and do not make it clear that many measures require
vigilance in use, sometimes more than the measure deserves.

35
2.5 Integration of Aesthetics & Performance

Among building performance and aesthetics there is one vital variable that is controlled
(shaped) by these two design factors, and that is the comfort of building occupants. Hausladen,
de Saldanha, & Liedl (2006) say:
“As people nowadays spend most of their time in rooms and hardly any time
outdoors, room climate has become increasingly important to our feeling of well-
being.”

And, Eisele & Kloft (2003) write:

“Comfort is measured in dynamic quantities, which must take the frequency and
type of activity as well as the location of the occupants into consideration.”

These authors suggest that it is essential to provide the building occupants with a healthy
work environment which can help with their wellbeing and productivity.

From an environmental perspective, some may argue that the energy/thermal performance of a
building is more important than its aesthetics. The following source supports this argument:
“Do you want to save serious energy and serious money? That’s easy. Use less
glass. Windows and curtain walls are the most expensive component in a building
and provide the worst energy performance.” (Lstiburek 2008).

However, there are many people arguing that from a building design of view, aesthetics is very
important. Hausladen, de Saldanha, & Liedl (2006) say:
“In additional to acting as a skin, façades have considerable influence on the
external appearance of a building. They can be used to provide information or
project an image to the public.”

Simulations tools can estimate the overall performance of a building design to guide designers
in the direction of maximising energy efficiency; testing various solutions to determine the best
design (in terms of energy performance).
“Building design should be no longer merely dominated by aesthetic and functional
considerations. Environmental performance based concern needs to be considered
at the planning stage, which can help to deliver valuable information on the viability
of a design approach.” (Pollock, Roderick, et al 2009).

The envelope of the building should assist the mechanical systems and work together as one to
provide a comfortable environment for the occupants. In order to be energy efficient, it should
never come down to full reliance on the mechanical systems.
“The aesthetic appearance of the exterior envelope need not conflict with the other
performance requirements, but as is the case with other performance requirements,
aesthetic considerations should not be allowed to predominate over the
achievement of other requirements.” (Persily 1993).

36
Early design tools enable a real balance to be achieved – not just a strong aesthetic gesture
and estimation of performance. Early design tools permit the integration of design, aesthetics
and performance. For example, a certain aesthetic could be achieved through an alternative
construction method that would benefit the building performance as well. The desired building
appearance and performance can be simulated by early design tools to assist practitioners on
making decisions, Flanagan & Norman (1983) wrote:
“A decision is being made to acquire assets that are intended to last and to be used
for a number of years. These assets will commit the owner or user not only to initial
capital costs, but also to subsequent running costs, day to day operating, cleaning
and maintenance costs, and periodic repair or replacement costs.”

Figures 19 & 20 show two unique (glass) buildings that speak for themselves through their
exterior aesthetics. They are examples of rhetoric building design about strong aesthetics
integrated with performance.

Figure 19 – Exterior view of the Mode Gakuen Cocoon Tower in Tokyo, Japan
Photograph © Tange Associates (Wood, 2010)
Figure 20 – Exterior view of the Mode Gakuen Spiral Towers in Nagoya-shi, Japan
Photograph © Nikken Sekkei Ltd (Wood, 2010)

37
The Cocoon Tower in Figure 19 is 50 stories (204m high) with its primary use being a school
and commercial retail. Tim Johnson (NBBJ) quoted:
“This is a great example of using a rich program to create a wide variety of spaces
that take advantage of natural light and inspiring views.” (Wood 2010).

Wood (2010) mentioned:

“The tower is designed specifically with the environment in mind. This includes a
cogeneration system, installed within the building, that produces about 40% of the
structure’s power and thermal energy. “

Also, “The elliptic shape allows for even distribution of sunlight, thereby limiting heat
radiation to the surrounding area.”

The Spiral Towers in Figure 20 is 36 stories (170m high). Its primary use is school with some
retail as well. Gordon Gill, Adrian Smith + Gordon Gill Architecture quoted:
“Finally we see a rotated tower scheme that has real integrity. It is well planned and
creatively varied in its spatial experiences.” (Wood 2010).

Wood (2010) says:

“The unique design of the three wings of the tower, twisted in helical form, appear
to change shape when viewed from different angles, giving an elegant yet dynamic
impression.”

Also, “Double-glazed windows and air-flow windows are employed to reduce heat
loads created by the sun around the perimeter zone.”

Wood (2010) noted that both of these buildings maintained their architectural integrity and
utilised different methods to help reduce energy consumption.

In summary, the factors of aesthetics and performance are just as important as each other in a
building design. Essentially, they are both part of the vital formula for designing a good
commercial building. Practitioners should express their architecture through performance of
design and through performance of design the aesthetics will emerge; if that is achieved,
comfort of occupants will automatically follow. However, if the aesthetics overshadows the
building performance, it is inevitable that the mechanical systems will be responsible for dealing
with the environmental factors such as excessive heat gain/loss throughout the year.

38
3 METHODOLOGY

In order to improve the relevance and consistency of the overall research method, this research
methodology consists of three separate individual research methods. The purpose is to
examine and explore the capability of the performance sketch early design tool concept. The
three methods used are: to develop a use-case study using COMFEN (LBNL 2010) as an
example; and then to explore whether a performance sketch tool can predict energy
performance estimations in a plausible manner; and finally to explore the impact on design
predictions of different levels of complexity of the performance sketch.

3.1 Overall Hypotheses

The following are the hypotheses for this research project:

• Performance sketch tool COMFEN can predict plausible energy performance
estimations that are consistent with results derived from detailed design tool
EnergyPlus.
• Use-case studies of COMFEN with practitioners will provide an evaluation of the
concept of the Performance Sketch concept resulting in a definition of the
properties of an ‘ideal’ early design tool.

The methodology examines these hypotheses by testing the single-zone (simulation) method of
COMFEN to see whether its predictions are plausible for energy analyses of commercial
building designs from an early design perspective.

For some forms of early design tool there are such simplifications that determining the reason
for implausible performance predictions is tricky. In this research, the advantage is the
underlying calculation engine for the early design tool is a simulation engine and the same
simulator (EnergyPlus) can be used to test consistency with detailed design predictions. The
differences revealed are thus most likely to be about the (performance) sketch – the
simplification process.

Ultimately, the results will demonstrate the potentials of COMFEN as a performance sketch tool
and identify areas that require improvement. It could eventually benefit many practitioners in the
building industry and help reduce energy consumption in heavily glazed commercial buildings.
The three-method approach used for this research is explained in the next section; these three
methods interlock together as one with the main focus on assessing the performance sketch
concept.

39
3.2 Methodology Overview

The research methodology used the following three research methods:

1. Use-case study of a Performance Sketch Tool with Practitioners

Seek feedback from a number of practitioners in the building industry on the subject of
COMFEN – the performance sketch tool; via interviews, the practitioners’ feedback are
‘use-cases’ (Haumer 2004) in this research.

2. Consistency Test of a Performance Sketch Tool

Testing of COMFEN example façades under contrasting climates to examine the

consistency between its energy predictions and results derived from the
EnergyPlus detailed model.

3. Evaluating the Adequacy of the Sketch within a Performance Sketch Tool

Explore practical application in COMFEN of users’ ideas where these are not already in
COMFEN (i.e. testing of some suggestions from the use-cases in EnergyPlus).

All three research methods are relevant in terms of improving the energy predictions by a
performance sketch early design tool. The outcome of each section is described in the results
chapter.

40
3.3 An Example of a Performance Sketch Early Design Tool:
COMFEN 3.0

COMFEN was selected as an example of a performance sketch tool for this research. This early
design application matches the performance sketch concept. Using a performance sketch tool
like COMFEN can be very helpful during the early design stage where very little design
detail/information is available.

COMFEN uses the sophisticated detailed design performance analysis program (EnergyPlus) in
a simpler more constrained – sketch-like – manner. This comes with the advantage of ensuring
that the sketch analysis calculation engine improves at the same rate as the detailed design
tool; and, it ensures that the performance information generated in COMFEN can be duplicated
and re-used as the model and the design itself get more sophisticated during the building
project.

The following statement is the definition/intention of the COMFEN software from its developer
(http://windows.lbl.gov/software/comfen/3/index.html):
“Lawrence Berkeley National Laboratory’s (LBNL) software – COMFEN
(COMmercial FENestration) is a tool designed to support the systematic evaluation
of alternative fenestration systems for project-specific commercial building
applications. COMFEN provides a simplified user interface that focuses attention on
key variables in fenestration design. Under the hood is EnergyPlus, a sophisticated
analysis engine that dynamically simulates the effects of these key fenestration
variables on energy consumption, peak energy demand, and thermal and visual
comfort.”

Figure 21 shows the typical COMFEN simulation interface. It displays the fenestration analyses
in simple (sketch) graphical form; it is capable of comparing multiple designs at once.

Figure 21 – COMFEN 3.0 simulation interface

41
COMFEN is an early design interface to the EnergyPlus detailed simulation engine. It provides
the user with a performance sketch interface which includes many vital simulation functions
required for an early design analysis.

COMFEN provides the user with a simplified diagram and energy performance graphs (i.e.
energy distribution) of each façade design option. It simulates one façade orientation and one
zone at a time for each fenestration scenario. It can simulate façades facing any desired
orientation. COMFEN 3.0 is capable of the following analyses through its graphical interface:
• Energy use distribution of HVAC, electric lighting and interior equipment loads
• Peak energy estimation
• Carbon dioxide (CO2) emission of each design scenario
• Daylight predictions (annual average illuminance)
• Discomfort glare index analyses
• Thermal comfort predictions – Predicted Percentage of Dissatisfied (PPD)
• Tabular data (percentage difference between base design and other scenarios)

COMFEN’s speed of calculation is significant; quick and efficient due to its single-zone
calculation method. It is a performance sketch tool that allows the user to assess fenestration
designs quickly without detailed design information.

This research focuses on the Energy Use Index (EUI) of the façade design derived from
COMFEN and EnergyPlus (i.e. energy end-use of electric lighting, interior equipment & HVAC)
for each façade design tested. The reason for this focus is that the research is about
determining the consistency of energy performance (advice/feedback) between early and
detailed design tools.

42
The following figures show the edit scenario windows within COMFEN where the building
project can be altered through elevation, plan and section views.

Figure 22 – COMFEN fenestration editing view in elevation

Figure 23 – COMFEN fenestration editing view in plan

Figure 24 – COMFEN fenestration editing view in section

43
Elevation view (Figure 22) allows the user to alter the fenestration design. For example, the
user can quickly sketch/assess various window sizes, glazing types, adding shading devices
and fins/overhangs. It facilitates the comparison of various combinations of glazing and shading
device to identify the best design concept(s) in a particular climate. The plan and section views
(Figures 23 & 24) provide a visualisation of the fenestration design (i.e. visual reference for the
location of windows/shading devices). Below each view, there are (white) boxes for simulation
input entry. These input options are just capabilities of the detailed design tool EnergyPlus.
COMFEN capabilities are made available for sketch/early design assessment purposes. The
user can alter the fenestration geometry, floor area, lighting & equipment loads (Watts per metre
square) and number of occupants. Also, the user can select the compass orientation they desire
to simulate and the type of lighting/glare control settings. The user can record the description of
each sketch, therefore different scenarios can be easily differentiated and documented.

There are built-in example files within COMFEN that help the user to learn the software; these
are examples of fenestration design scenarios/projects. Also, there are libraries of glazing type
and shading device system which allows the user to create their own fenestration designs (i.e.
combinations of glazing and shading device). The purpose of these example files is to help the
user understand the sketch application and its simulation inputs/outputs. When they become
more familiar with the software, the user can quickly duplicate and alter these example files to
form combinations of fenestration design to suit a specified project brief.

In summary, architects and engineers can pin-point the best fenestration concept designs with
COMFEN multiple façade comparison capability. The lessons learnt from the early design
predictions calculated by COMFEN can assist practitioners in making design decisions from day
one of the building project. Hitchcock, Mitchell, Yazdanian, Lee, & Huizenga (2008) state:
• The overall objective of COMFEN is to promote the design and deployment of
high performance fenestration systems by making complex simulation
comparisons of alternative fenestration design choices accessible to a wide
audience of users.

• The primary audience of COMFEN is architects during early design. A second

primary audience of COMFEN is fenestration manufacturers.

• The main objective of COMFEN in this respect is to analyse new complex

fenestration products in an unbiased manner so that those products that enhance
the overall performance of fenestration systems can be promoted to consumers.

COMFEN is a performance sketch tool specifically designed to assess ‘commercial fenestration’

during the early stages of design.

44
3.4 An Example of a Detailed Design Tool: EnergyPlus 4.0

EnergyPlus is a comprehensive (multi-zone) building energy simulation engine. The desired

building design determined by an early design tool can be carried into EnergyPlus for further
and more complex simulations (i.e. during detailed design stage).
“EnergyPlus is a whole building energy simulation program that engineers,
architects and researchers use to model energy and water use in buildings.
Modelling the performance of a building with EnergyPlus enables building
professionals to optimise the building design to use less energy and water.” (US
DoE 2010).

EnergyPlus has the power to model (simulate) complex details within a building design (e.g.
finalized geometries, materials, mechanical systems & building schedules).

Figure 25 – EnergyPlus 4.0 idf-Editor

The detailed design stage for a building is the process before the actual construction begins. No
physical work can begin until every detail of the building design is examined and approved.
Areas that are specifically related to this research are HVAC and lighting energy predictions (i.e.
EUI predicted by COMFEN and EnergyPlus). All the EnergyPlus simulation models in used by
COMFEN and EnergyPlus are of ‘.idf’ file type.

The EnergyPlus ‘idf-Editor’ interface is displayed in a spreadsheet-like format as can be seen in

Figure 25. Hundreds of design variables can be entered by the user to the desired settings (e.g.
HVAC system, lighting loads, occupancy & building schedules etc). The idf-Editor does not have
a graphical representation of the simulation inputs (i.e. model geometry). However, there is a
Plug-in for SketchUp that makes EnergyPlus modelling much more user-friendly. The following
section explains why the OpenStudio Plug-in was used in this study.

45
EnergyPlus allows the designer to simulate and predict through hour by hour calculations for a
year how a building would perform when completed. Its link to COMFEN is a powerful
relationship (i.e. software compatibility) that allows practitioners to apply more (or less) detail to
a simulation model during any time throughout the design phase; it offers the advantage of
referring back to the performance sketch tool during the detailed design stage, when necessary,
secure in the knowledge that the calculation engine in both cases is EnergyPlus and thus the
calculation process will be consistent.

46
3.4.1 OpenStudio Plug-in

OpenStudio was used in this research because it is an essential visual integrity check on the
geometry modelled in EnergyPlus. Figure 25 shows the spreadsheet-like format in which
geometry and all other data describing a building is entered into EnergyPlus. The visual cross-
reference provided by OpenStudio would be a very useful check of the data as a visualizer of
the data. The advantage of OpenStudio is that the geometry can also be edited.

The following statement is the definition of the OpenStudio Plug-in from its developers
(http://apps1.eere.energy.gov/buildings/energyplus/openstudio.cfm):
“OpenStudio Plug-in allows you to use the standard SketchUp tools to create and
edit EnergyPlus zones and surfaces. You can explore your EnergyPlus input files
by using all of the native SketchUp 3D capabilities to view the geometry from any
vantage point, apply different rendering styles, and perform shadowing studies. The
plug-in allows you to mix EnergyPlus simulation content with decorative content
such as background images, landscaping, people, and architectural finish details –
all within the same SketchUp model.”

Figure 26 – OpenStudio within SketchUp

Figure 26 is an example of the building geometry of an EnergyPlus file viewed via the
OpenStudio Plug-in within Google SketchUp. In this research OpenStudio was mainly used for
the geometry modelling, minor changes to building envelope materials and shading devices.
This enabled a more consistent modelling of geometry within the idf-Editor, which can become
difficult when purely handled via numerical data entry. All inputs excluding geometry
coordinates were entered within the spreadsheet-like format of the idf-Editor. For example,
some of these details were: HVAC system, lighting loads, occupancy, and building schedules.

47
3.5 Introduction of a Performance Sketch Tool to Practitioners

The approach of the first method of this research is to present COMFEN to practitioners as an
example of a performance sketch tool. The selection of practitioners was from as wide a range
of backgrounds as possible. Their feedback on the concept of the Performance Sketch was
collated into a specification of their ‘ideal’ early design tool.

The intention was to show the potentials of COMFEN to practitioners and through structured
questioning to gather their opinions on the performance sketch tool idea. Feedback from
architects and engineers from New Zealand, Japan, and Taiwan and the USA was obtained.

The use-cases play a central role in the analysis of the Performance Sketch. The practitioners’
feedback on their early design needs was based on past and current project experiences.

Structured interviews were held with practitioners centred on a list of questions relating to
COMFEN as an example of a performance sketch early design tool. The use-cases were
intended to determine what worked and what additional capabilities were desired. The research
design allowed the possibility that if a consensus list of desirable design options could be
compiled for COMFEN then these design options could be tested within EnergyPlus and thus
be validated for future versions of COMFEN.

The ultimate objective is to have an early design tool that architects and engineers can trust and
use to deliver their design to a client. To do that the tool has to minimise its design limitations,
because project bids are often won or lost based on the building’s aesthetic; so having the
freedom to analyse a wide range of architectural concepts is essential.

In this research, an information sheet and consent form regarding this research were distributed
(emailed) to various architectural/engineering firms around the world. The practitioners selected
for the interviews were based on their personal interest/response to the performance sketch
idea. The practitioners interviewed were all new to the COMFEN software and they were asked
to provide feedback after an introduction to the software.

The use-cases illustrate a wide range of views stated by professionals from different
backgrounds, countries, cultures, and practices.

48
3.5.1 Interview Sample Size

Turner, Lewis, & Nielsen (2006) state that:

• most usability problems are detected with the first three to five subjects

• running additional subjects during the same test is unlikely to reveal new information

• return on investment (ROI) in usability testing is maximized when testing with small
groups using an iterative test-and-design methodology

These claims are determined in the studies by Virzi (1992), Nielsen & Landauer (1993) and
Lewis (1994). On the basis of these sources, it was decided that only a small number of
participants would be sufficient to carry out this research. The first five practitioners who
responded to the research information sheet were chosen for face-to-face interviews.
Sometimes more than one practitioner attended the use-case study interviews so the data
represents the views of 6 firms, but 9 people.

3.5.2 Delphi-like Review of Responses

The ‘Delphi’ method was incorporated into the interviews with practitioners. This allowed them
during the use-case visits to freely express their opinion on COMFEN as an example of a
performance sketch early design tool. Then the Delphi review process allowed for consultation
about the form and content of the notes taken, and to consider others’ views and contribute to
the development of a consensus view based on professional work experiences. Rowe & Wright
(1999) characterised the Delphi method with four key features:
1. Anonymity: allows the participants to freely express their opinions without undue
social pressures to conform from others in the group. Decisions are evaluated on
their merit, rather than who has proposed the idea.

2. Iteration: allows the participants to refine their views in light of the progress of the
group’s work from round to round.

3. Controlled feedback: informs the participants of the other participant’s

perspectives, and provides the opportunity for participants to clarify or change
their views.

4. Statistical aggregation of group response: allows for a quantitative analysis

and interpretation of data.

“The Delphi method is an iterative process used to collect and distill the judgments
of experts using a series of questionnaires interspersed with feedback. The
questionnaires are designed to focus on problems, opportunities, solutions, or
forecasts.” (Skulmoski, Hartman, & Krahn 2007).

Interview questions are stated in Section 4.1.1 in the Results Chapter.

49
3.5.3 Process of Interviewing Practitioners: Introduction of COMFEN

COMFEN was introduced to various practitioners as an example of a performance sketch early

design tool. The interview use-cases were from:
• 6 different architecture & engineering firms around the world; which includes a
USA interview – recorded in 2008 prior to the commencement of this thesis (Donn
2010)
• Total practitioners contributed in uses-cases = 9 male professionals
(5 architects & 4 engineers).

The interview process comprised four phases:

1. Research topic information sheet and consent form were approved by the ethics
committee of Victoria University of Wellington, New Zealand.

2. Information sheet and consent form were distributed by email to various

architectural/engineering firms around the world.

3. Practitioners (architects & engineers) in Wellington, Japan and Taiwan agreed to

take part in the research. Their feedback is the use-cases in this thesis.

4. Interview arrangements (visit selected practitioners)

Note: The participants were selected based on their personal interest in the Performance
Sketch Concept (i.e. not randomly selected).

There are always barriers when the research goal involves gathering feedback from working
professionals in the building industry. There are many factors that influence the participant on
taking part in the research such as: personal interest on the research; their availability of time;
the firm’s rules on design confidentiality.

All interview participants were new to the COMFEN software. They are male architects and
engineers in the building industry. They were encouraged to express their opinions on
COMFEN based purely on their own perspectives. APPENDIX A & APPENDIX B contain,
respectively, the information sheet sent to practitioners.

50
3.6 Consistency Test of a Performance Sketch Tool

This section describes the second method of this research. The purpose of this test was to
assess COMFEN as an example of a performance sketch early design tool, to see whether its
single-zone predictions are consistent with results derived from detailed multi-zone simulations.
It was assumed that for COMFEN to be viewed as a plausible early design analysis tool, the
predicted performance should be carried into the detailed design stage.

COMFEN is EnergyPlus, it only makes certain functions of EnergyPlus available for early design
purposes. The great advantage of this sketch tool is that more detail can be applied to a model
(as the design progresses) because of its relationship/compatibility to EnergyPlus. And it
provides the user with the capability/functionality of a complex simulation tool during the
preliminary design phase.

COMFEN (version 3.0) was specifically designed to analyse façade designs for single perimeter
zones only; the performance sketch used by COMFEN to make the calculations quick is that of
the single zone in a building. And detailed simulations in EnergyPlus (version 4.0) can examine
the entire building.

The methodology adopted for this test examines COMFEN with four of its built-in example
façade designs (various glazing types and shading combinations) under four climate zones. It
compares the COMFEN performance predictions with those from EnergyPlus when these
façade designs are applied to full scale buildings. The geometry selected for the full scale
buildings is the geometry of DoE’s small & large benchmark offices (US DoE 2010). They were
modelled under the same climates.

This test examines COMFEN as an example of a performance sketch tool by assessing the
consistency of single-zone predictions with results derived from multi-zone simulations in
EnergyPlus. The simulation analyses performed in this test are:
• COMFEN analysis of four façade designs (built-in examples see section 3.6.2).
• COMFEN façades incorporated to full scale DoE office buildings in EnergyPlus.

The full scale (multi-zone) buildings use the same default settings as COMFEN. The only
difference between the single-zone and multi-zone models is the scale (i.e. floor area).

If COMFEN’s single-zone method can produce energy results that are consistent with multi-
zone (detailed) simulations it would be beneficial to the users.

51
3.6.1 The Definition of Consistency

The building energy simulations focused on the Energy Use Index (EUI) of the façade design
derived from COMFEN and EnergyPlus. It was used to determine the ranking of energy
performance (electric lighting load, interior equipment load & HVAC load) for each façade
design tested. It is important to note that the significance of the simulation results lies within the
relative values (i.e. distribution of energy end-uses), not the absolute value of a fenestration
design’s total EUI.

The rankings of façade performance predictions in COMFEN and EnergyPlus determine the
consistency of single-zone calculation method to the multi-zone method. Energy performance
ranking was determined by fenestration energy predictions; façade design with the lowest total
2
EUI (kWh/m .yr) is ranked first.

The definition of consistency is:

• COMFEN (single-zone) and EnergyPlus (multi-zone) predictions of energy
performance will produce the same design advice/feedback (i.e. the ranking of the
façade ideas will be the same, as well as similarity in distribution of energy end-
uses).

Distribution of energy end-uses determines the consistency of performance sketch

advice/feedback. For example, while total energy use might be the same, the distribution of
energy end-uses has to be consistent between single-zone and multi-zone models. In order the
design advice/feedback to be consistent there should be consistency in HVAC & Lighting
energy demand.

52
3.6.2 Consistency Test: Fenestration Design Scenarios

In the first stage of this test, four generic fenestration designs were simulated in COMFEN. The
following are the chosen COMFEN façade designs:
A. Double clear Low-E glazing with exterior venetian blinds (permanently fixed) at 45°
B. Double clear Low-E glazing with between venetian blinds (permanently fixed) at 45°
C. Double clear Low-E glazing with interior venetian blinds (permanently fixed) at 45°
D. Single clear glazing

Figure 27 shows the section details of the façade designs modelled in COMFEN (for the full
description of the designs, refer to APPENDIX I – Façade Constructions & Material Properties).

Figure 27 – Façade designs simulated in COMFEN & EnergyPlus

The reason for these fenestration selections is to create façades with the potential for radical
differences in energy performance. The façade designs are similar to each other apart from the
position of Venetian blinds, which can lead to a simple interpretation of the performance
differences. In Figure 27, the left of each fenestration option represents the exterior
environment and the right is the interior. The letter code of each façade design (Figure 27) links
to graphs in the results chapter.

These following locations were chosen for this test:

• Wellington, New Zealand – mixed, marine climate zone (IWEC.epw)
• Melbourne, Australia – warm, marine climate zone (IWEC.epw)
• Minneapolis-St Paul, USA – cold, humid climate zone (TMY2.epw)
• Bombay, India – mixed, dry climate zone (IWEC.epw)

Note: TMY2 weather file was from NREL and IWEC files were from ASHRAE (for each location
the same weather file was used in both COMFEN and EnergyPlus simulations).

53
There is a possibility that design predictions are consistent in one climate zone and inconsistent
in another. Therefore a range of different locations were selected. Contrasting climate zones
were selected for testing in order to assess particular single-zone/multi-zone design
issues/properties; examine how COMFEN fenestration designs perform in the best/worst
scenarios.

In COMFEN, simulations are performed in single zones only, so there is a zone for each
façade orientation (Figure 28). The following figure is a default COMFEN zone which has
2
a total floor area of 37m². The R-value of COMFEN default walls = 2.29 m -K/W; Window-
to-Wall Ratio (WWR) was set to 58% in all models.

Figure 28 – Typical COMFEN single-zone model and its dimensions

For each fenestration design, four orientations (4 zones) were simulated to calculate the
average of the north, south, east and west-facing zones of a whole building. The average
st th
fenestration energy predictions were used to develop an energy performance ranking (1 ~4 )
because the average of four (COMFEN) perimeter zones represents an entire building for each
façade design. The ranking was developed to show which façade design is the most energy
efficient. This was intended to mirror the way in which a designer would use a performance
sketch to determine the optimum façade for a building.

54
3.6.3 Fenestration Design Scenarios Integrated into DoE Benchmark Models

The façade designs used in the COMFEN simulations were integrated into the US Department
of Energy’s (DoE) small and large benchmark offices (US DoE 2010) under the same climate
zones as the COMFEN designs.

In order to ensure the EUI comparison between COMFEN and EnergyPlus predictions was
focused only on the two programs’ ranking of the façade designs the models in each program
had a consistent Window-to-Wall Ratio. All models in this test had a WWR of 58% which
originated from the DoE large office model.

st th
The façades’ energy performance was ranked from 1 to 4 in both programs. Figures 29 & 30
are SketchUp’s representation of the benchmark offices in three dimensional (X-ray views)
forms within OpenStudio; dividing walls in Figures 29 & 30 show the separation of zones in
each office building.

Figure 29 – Geometry of DoE small benchmark office

Figure 30 – Geometry of DoE large benchmark office

55
In the large office, EnergyPlus divides up floors into top, middle and bottom for its calculation
because energy use will vary in each floor level. The method of energy calculation consists of
top, bottom and basement floors (each multiplied by 1) plus middle floors (multiplied by 10) to
represent the 13 storey benchmark large office. The small office is simply displayed as a single
storey building in OpenStudio.

The benchmark offices consist of multiple zones, where each floor is made up of a single core
zone and four perimeter zones. A core zone (internal space) does not have as much daylight
exposure (solar gain) compared to perimeter zones. Figure 31 shows the plan of the zones in
the small office. The 5 zones have the same arrangement as the 5 zones on each floor of the
large office. The principal differences between small and large office are the scale of the
building and its zones (CZ = core zone & PZ = perimeter zone). Figure 32 shows the floor layout
of the large benchmark office.

Figure 31 – Floor layout of perimeter zones and core zone in the small benchmark office

Figure 32 – Floor layout of perimeter zones and core zone in the large benchmark office

The floor area of the DoE benchmark offices simulated in EnergyPlus is:
• Small office: 510m² - total of 1 storey (attic excluded in EUI) Figure 29 & Figure 31
• Large office: 46,320m² - total of 13 stories (basement included) Figure 30 & Figure 32

56
The difference in floor area between small and large offices is immense and it is likely to be
significant in any energy performance calculations. Figure 32 shows that the large office’s
perimeter zones have a depth which is much smaller as a proportion of the whole floor area
compared to the perimeter zones in the small office. The distribution of core zone(s) and
perimeter zones could have a significant influence on the energy performance predictions as
the energy needs of a zone in contact with the outside climate are very different than the energy
needs of a zone almost completely isolated from the outside. The reason for assessing these
two different office models is to examine whether this scale difference in a building makes a
difference in terms of overall energy performance and façade ranking.

Table 1 below shows how the floor area percentage of perimeter zones versus core zone(s)
varies in the DoE offices.

DoE Office m² % of Perimeter Zones m² % of Core Zone(s) Total m² % of building

Small 71% (362m²) 29% (148m²) 100% (510m²)
Large 29% (13,433m²) 71% (32,887m²) 100% (46,320m²)

Table 1 – Floor area percentage of perimeter zones versus core zones

The energy use of the perimeter zones varies depending on their façade orientations. The core
zone is not directly connected to daylight therefore it has a higher electric lighting energy
demand than perimeter zones. One of the principal focuses of this research is therefore
examination of the influence on the consistency of energy use predictions of the COMFEN
model with no core zones when compared to full scale (multi-zone) models.

The integration of COMFEN façade designs into the DoE offices created four different buildings
in each office category because of the four fenestration designs examined in this test. Each
façade design represents an individual model with identical fenestration design on all
orientations (e.g. Small office with design A, Small office with design B, Large office with design
A etc). A total of two building types and four façades, therefore eight multi-zone buildings were
tested across four climate zones. The results are compared with COMFEN predictions to see if
the single-zone design messages are consistent.

57
3.6.4 Simulation Settings

The following are the general simulation settings applied to the COMFEN and EnergyPlus
models:
Location: according to weather file
North Axis: 0 degrees
Terrain: suburbs (COMFEN default)
Daylight control: daylight used when working surface exceeds 538 Lux which reduces the
electric lighting load during that period. Electric lighting is switched on when working surface
drops below 538 Lux (538 Lux is derived from COMFEN which is approximately 50fc)
Shading Control Type: always on
HVAC: fully HVAC conditioned zones
Fenestration construction: see Figure 27 & APPENDIX I
2
Building envelope: COMFEN default construction settings (Wall R-value = 2.29 m -K/W)

Table 2 – General simulation settings used in the consistent test of COMFEN

The type of HVAC system and targeted standards used in the simulation models are:
• PSZ-AC: DoE’s Packaged Single Zone – Air Conditioner HVAC system
• Targeted Standard: ASHRAE Standard 90.1-2004 & 62-1999 (energy standard
for buildings except low-rise residential buildings)

Table 3 shows the detailed settings used in COMFEN and EnergyPlus models:
Modelling setting COMFEN & EP models
Window-to-Wall Ratio (%) 58
2
Occupancy (people/m ) 0.03
2
Interior Lighting (W/m ) 13.5
2
Interior Equipment (W/m ) 8.1
Infiltration (ACH) 0.3
HVAC System PSZ-AC
Building Schedules COMFEN Small office
Targeted Standard ASHRAE 90.1-2004

Table 3 – Further detailed simulation settings in COMFEN and DoE offices

As many simulation settings as possible were made consistent between the COMFEN and
EnergyPlus models to ensure that any observed differences in energy use prediction would
focus on the modelling approach – the single zone of COMFEN versus the multi-zone of
EnergyPlus.

58
3.7 Evaluating the Adequacy of the Sketch within a Performance
Sketch Tool

This section describes the third method of this research. It examines the use-cases (interview
feedback) by exploring the complexity of COMFEN as an example of a performance sketch tool.
The main suggestions regarding COMFEN improvements are (see Table 7 on page 69):
• the ability to examine/understand the performance of adjacent zones (i.e. multi-zones)
• the ability to simulate ‘Natural Ventilation’

The simple 1-zone simulation approach of COMFEN has its benefits. These are: quick
calculations and the ability to compare multiple façade designs at once. However, its current
application does not allow the user to assess designs with natural ventilation, which rules out
this energy saving opportunity. Also, focusing on perimeter zones of a building design could be
misleading because core zones could have a huge effect on the overall energy consumption.

The goal of this method was to use EnergyPlus to “mock up” a COMFEN run to examine the
design options reported as desired during the use-cases. This method relies on the fact that
EnergyPlus is the calculation engine for COMFEN. If a COMFEN single-zone model were to be
simulated in EnergyPlus the results would be the same. This mock-up process examined how
one might create a performance sketch early design tool based upon a more complex model
than the single perimeter zone of COMFEN.

3.7.1 Adding Complexity to the Performance Sketch Tool

Two particular aspects of Performance Sketch complexity were modelled with full version of
EnergyPlus; they were:
Is a multi-zone sketch model different to a generic COMFEN single-zone model? Does it
produce the same design lessons/feedback?
The multi-zone sketch is assumed to be a more accurate method (i.e. modelling of
associated spaces within a building design) than the single-zone approach when
modelling building energy performance. The purpose of this analysis is to examine
whether it is possible to calculate results similar to multi-zone simulations based on
COMFEN’s (single-zone) method. Multi-zone tests were performed in EnergyPlus to
explore the multi-zone design option. Also, to prove that single-zone (simulation) is
enough from an early design perspective.

Could COMFEN run a single-zone/multi-zone natural ventilation model?

The purpose of this analysis is to assess the design option of natural ventilation with
single and multi-zone models in EnergyPlus. Tests were performed in EnergyPlus
because the current COMFEN interface cannot simulate natural ventilation. Potentially,
natural ventilation could be incorporated into future versions (depending on verification
of results).

59
This analysis addressed two simulation types for each category of modelling method:
• HAVC Only models (current intention of COMFEN)
• Mixed Mode models (desired feature of Natural Ventilation incorporated with HVAC)

The following are the categories of modelling method explored in this analysis:
• COMFEN model: Single-zone in COMFEN. Consists of north & south facing
zones simulated separately (for HVAC Only model).
• 1-zone model: Single zones in EnergyPlus to assess Natural Ventilation option
(for Mixed Mode model: north & south zones simulated separately).
• 2-zone model: Consists of 2 zones in EP (north & south zones).
• 3-zone model: Consists of north, south and core zones in EP. The 3-zone model
is the equivalent of three generic COMFEN single zones, each zone with the
same width, length and height (3m).

Note: all models have the same height, width, and length with identical glazing percentage. The
only difference is the zone layout and connections from zone to zone. It is significant to have the
same dimensions across the modelling method categories because they are likely to have a
large effect on the final calculated performance.

Figure 33 – Plan view: modelling categories

Figure 33 shows the zone layouts modelled. On the left is a ‘single-zone’ model – the North and
the South facing zones are modelled in separate COMFEN like runs and the total energy use is
the sum of the two zone energy uses; in the centre is a 2-zone model – one model comprising
two zones connected by an internal wall; on the right is a 3-zone model where the interior zone
has no fenestration connection to the outdoor climate. The floor areas are the same in all
models at 112m² (equivalent of 3 standard COMFEN zones).

60
The only difference between the mock up COMFEN model and the real one is that you cannot
model the Natural Ventilation and multi-zone use-case suggestions through the current
COMFEN interface.

The 2-zone test questions whether 1-zone is enough to assess a fenestration design. The 3-
zone model illustrates the effect of an interior (core) zone. The results will determine whether
the simplest approach to creating a performance sketch (1-zone calculations) is adequate for
early design analyses.

2
The energy use index (EUI: kWh/m .yr) of each modelling category was assessed to determine
the EUI percentage difference between the calculation methods (i.e. single-zone vs. 2-zone
model & single-zone vs. 3-zone model).

61
3.7.1.1 Simulation Settings:

The following simulation assumptions are kept consistent throughout the modelling categories;
these settings are significant because they could have a huge effect on the calculated
performance of each modelling category. Therefore the simulation settings need to be: a) the
same for every simulation; b) in ranges that make the building convincing models of actual
commercial building use (i.e. heavily glazed buildings illustrated in Section 1.3).

The following are the general simulation settings applied to models within this analysis:
Location: Wellington, New Zealand
North Axis: 0 degrees
Terrain: suburbs (COMFEN default)
Daylight control: daylight used when working surface exceeds 538 Lux which reduces the
electric lighting load during that period. Electric lighting is switched on when working surface
drops below 538 Lux (538 Lux is derived from COMFEN which is approximately 50fc)
HVAC only models: fully HVAC conditioned zones
Mixed Mode models: natural ventilation used when indoor temperature exceeds 25°C and
switched off when interior temperature drops below 18°C
Fenestration construction: 6mm single clear glazing (see APPENDIX I)
2
Building envelope: COMFEN default construction settings (Wall R-value = 2.29 m -K/W)

Table 4 – General simulation settings for 1~3-zone models

The following table shows detailed simulation settings derived from a COMFEN curtain wall
example:
Modelling settings 1~3-zone models
Single Glazed Façade (mm) 6
Glazing Percentage (%) North & South façades 97% (relate to buildings in Section 1.3)
2
Occupancy (people/m ) 0.03
2
Interior Lighting (W/m ) 13.5
2
Interior Equipment (W/m ) 8.1
Infiltration (ACH) 0.3
PSZ-AC: Packaged Single Zone – Air
HVAC System
Conditioner HVAC system
Building Schedules COMFEN Small office
Targeted Standard ASHRAE 90.1-2004

Table 5 – Detailed simulation settings for 1~3-zone models

62
3.7.2 Individual Zone Comparison: Results of Adding Complexity to the Performance
Sketch Tool

The purpose of this test is to examine each individual zone simulated using a range of different,
more complex performance sketches. The goal is to determine whether a multi-zone approach
to creating the performance sketch is better or worse than the current one. The distribution of
2
energy use (kWh/m .yr) between each individual zone within the models is examined. The
analysis illustrates the amount of energy consumed by each zone in relation to its façade
orientation. This includes the five major energy components: Heating, Cooling, Ventilation,
Interior Lighting, and Interior Equipment. Figure 33 shows the zone layout of each modelling
category.

The question asked is: Is a single-zone sufficient for early design analyses? Is it capable of
producing similar design lessons (i.e. distribution of HVAC & lighting loads) to multi-zone
models?

The distribution of energy use between each zone for a fenestration design is as critical as the
EUI ranking. Consistent early design advice/feedback is not just predictions of a consistent
ranking of the EUI but also an indication of where the energy performance varies. The designer
could still be lead to an erroneous design decision if the distribution of energy end-uses is
inconsistent with what will happen in reality. For example, a designer could view the energy
performance of a three zone ‘sketch’ and note that the zone facing the noonday sun has the
tendency to overheat (i.e. high cooling demand). Their response could be radically different if
they viewed outputs from a one zone model which will not overheat even though it has the
same window and floor area as the three zone model. The following is how the percentage of
energy use index (EUI %) is determined in each modelling category:

• COMFEN model: Total EUI % = North zone EUI % + South zone EUI %
• 2-zone model: Total EUI % = North zone EUI % + South zone EUI %
• 3-zone model: Total EUI % = North Zone EUI % + Core zone EUI % + South Zone EUI %

2 2
E.g. North zone total EUI % = total kWh/m .yr (EUI) of north zone / total kWh/m .yr (EUI) of
2 2
model. South zone total EUI % = total kWh/m .yr of south zone / total kWh/m .yr of model.

Note: Total EUI = Total (annual) energy consumption of all zones within a model.

63
3.7.3 A Consistency Test for the Study of Adding Complexity to the Performance
Sketch Tool

Some of the concerns of the use-case interviewees related to the need for natural ventilation
and multi-zone simulation during early design analyses. These cannot be currently modelled
with COMFEN. They can be modelled with EnergyPlus. EnergyPlus was therefore used to mock
up the operation of COMFEN with these features. This was a test of adding a different sort of
complexity to the performance sketch tool.

The analysis is similar to the consistency test for COMFEN. The purpose of this analysis is to
examine single-zone and multi-zone models with natural ventilation incorporated. This was to
determine whether it is possible for single-zone models to produce the same design
advice/feedback as multi-zone models.

In this test two fenestration designs (Figure 34) were simulated to see if the single-zone energy
predictions are consistent with the multi-zone results (2-zone and 3-zone models). Therefore,
each fenestration design is simulated in all modelling categories; Mixed Mode is the only mode
assessed in this test (i.e. HVAC & Natural Ventilation).

Two fenestration designs were selected in order to explore the potential differences in energy
performance in the single-zone and multi-zone models with Mixed Mode incorporated. The
following are façade design details for this test (see ). Figure 33 shows the zone layout of each
modelling category.

Figure 34 – Façade designs simulated in consistency test

A. Double clear Low-E glazing with exterior venetian blinds (permanently fixed) at 45°
B. Double clear Low-E glazing with between venetian blinds (permanently fixed) at 45°

64
 2 st
Fenestration design with the lowest energy consumption (kWh/m .yr) was ranked 1 in
2 2
COMFEN and EnergyPlus (e.g. 100 kWh/m .yr is ranked above 150 kWh/m .yr etc).

Similar to the consistency test for COMFEN, consistency is achieved if single-zone and multi-
zone energy predictions are the same design advice/feedback. For example, energy
performance of a three zone ‘sketch’ has to be similar to a one zone model for the design
advice/feedback to be consistent; therefore the calculated results need to be consistent in: a)
ranking; b) distribution of energy-end-uses.

3.7.3.1 Simulation Settings

Simulation settings are kept the same throughout this consistency test between single-zone and
multi-zone models; these settings are significant because they could have a huge effect on the
calculated performance of each modelling category. The results illustrate the difference in
energy performance between fenestration designs tested in single-zone and multi-zone models.
Refer to Tables 4 & 5 for simulation settings.

65
66
4 RESULTS

This chapter reports the results of the three research methods described in the previous
chapter. It deals in turn with the interview use-case studies; the consistency test; and the study
of the complexity of a performance sketch tool.

4.1 Use-cases: Introduction of a Performance Sketch Tool to

Practitioners

This section analyses each practitioner’s interview feedback. Practitioners’ anonymity was
maintained by reporting their responses under labels A through F. The appendix chapter lists
their occupation within the firm, as architect or engineer. In relation to COMFEN as an example
of the performance sketch tool, initial feedback from each interview was also recorded in
APPENDICES C to H.

4.1.1 Interview Questions

The following questions were used to structure the interview with each practitioner after the
introduction of the COMFEN software:
1. What do you think of COMFEN?
2. Was there enough information in the graphical results?
3. Areas that need improvement?
4. Any further comments?

A practitioner was asked to provided further feedback if their comments challenged the
performance sketch tool idea. This was to determine not just the challenge, but also their
rationale for their feedback. Gathering further feedback on what would be the ideal ‘sketch’
design tool (i.e. its capability and usability) will help improve the relevance of COMFEN
predictions.

nd
The following questions were asked in the 2 round of feedback (if required):
1. If you were to design an early design tool (performance sketch tool), what would you
incorporate into the tool? How complex is too complex?
2. As a practitioner (keep in mind the performance sketch tool idea), what would be a good
early design tool to you? What is useful in very early sketch design?
3. After a better understanding of the performance sketch tool idea, does the feedback on
COMFEN change in anyway? Or the original interview feedback still stands?

The next few pages summarise the analysis of the use-cases and the subsequent feedback.
Note: If no further feedback was provided by the practitioner, the first round of feedback was
taken to be their final statement on COMFEN as a performance sketch tool.

67
4.1.2 Overall Analysis of Conclusions: Use-cases

Tables 6 & 7 summarise the practitioners’ views of the advantages of COMFEN (Table 6) and
the desired features for a performance sketch tool (Table 7). The reference column in each
table links to APPENDICES C~H which contain initial feedback recordings and individual
analysis of each use-case.

The following table is a summary of COMFEN advantages:

Advantages of COMFEN Reference to Appendix

Ability to compare multiple façade designs simultaneously
D, E, F, H
which are supported by good graphical information
The software is simple, quick and user-friendly from an early
C, E, F, G
design perspective
The program is compatible with EnergyPlus (detailed design
E
tool) which can carry the design into the detailed design stage
Good for communicating with the client through its graphical
G
interface (e.g. energy performance of the design)
It would be useful for window manufacturers (i.e. input various
G
products into the software)

Table 6 – Advantages of COMFEN

The results in Table 6 show the main advantages of COMFEN quoted by practitioners; these
were:
• The ability to compare multiple façade designs simultaneously which are
supported by good graphical information (4/6 use-cases).
• The software is simple, quick and user-friendly from an early design perspective
(4/6 use-cases).

Overall, practitioners had similar comments on COMFEN as a performance sketch tool. The
majority of the practitioners believe these advantages are great from a performance sketch
design perspective.

68
The following table is a summary drawn from practitioner feedback of the desired features for
future versions of the performance sketch tool:

Desired Features Reference to Appendix

Double façade design option C, D, E, G, H
Sketch/Simulate what you want
E.g. ability to sketch desired geometries and draw in shading C, D, F, H
elements rather than just define dimensions.
Aesthetics visualisation: realistic sketch of design
scenarios
C, D, F, H
E.g. colour of glass, fins, shading devices & orbit around the
3D model.
Annual cost analysis
E.g. option to plot in cost of energy per kilowatt-hour, life cycle D, E, F, G
cost calculation & material costs.
Multi-zone simulations
E.g. to assess the whole building, air/heat movement between
D, E, F
the perimeter zones & core zones – could influence the design
predictions dramatically.
Natural ventilation option
E.g. mixed mode option, control over size & schedule of C, D, F
openings.
Atrium design option D, E, G
Automatically compare to desired standard(s) C.2, H
Importing complex architectural geometry into the software
D, F
E.g. import drawings from Revit & other CAD programs.
Simulating external shading and sun path study
C, H
E.g. neighbouring buildings & trees.
Apply materials to opaque areas
D, G
E.g. walls/ceiling/floor.
HVAC system control options
E.g. summer & winter - seasonal efficiency, temperature
C.2, F
settings and more detail/information on the type of system –
Package Single Zone-AC.
Export models to other common detailed design tools (not
just EnergyPlus) F
I.e. compatible file format.
Control over internal gains
C.2
E.g. option to enter desired values.

Automatically convert simulation outputs to desired unit

2 2 C.2
E.g. kWh/m .yr & W/m etc.
Automatically include a small image of each façade option
on charts as well C.2
I.e. not just a number reference to design scenario.
Automatically produce charts comparing designs that
C.2
allow the user to bring into MS Excel
Light (Lux) measurement on ceiling
E.g. light bouncing off shading device – opportunity to use D
daylight with electric lights.
Solar gain & heat gain analyses H

Table 7 – Desired design features in future versions

69
Table 7 shows all the desired feedback reported during the use-cases. The features that were
most commonly referred to in the 6 use-cases, and thus might be viewed as of highest priority,
in sketch performance tools development beyond what COMFEN provides were:
• Double façade design option (5/6 use-cases).
• Sketch/Simulate what you want (4/6 use-cases).
• Aesthetics visualisation: realistic sketch of design scenarios (4/6 use-cases).
• Annual cost analysis (4/6 use-cases).
• Multi-zone simulations (3/6 use-cases).
• Natural ventilation option (3/6 use-cases).
• Atrium design option (3/6 use-cases).

Having the ability to sketch freely seems to be the essential feature proposed by the
practitioners. In order for the user to explore/express their architectural ideas, sketch limitations
must be kept at a minimum. Multi-zone simulations, natural ventilation, double façade and
atrium design features could be perceived as part of the ‘sketch/simulate what you want’
desired feature category. These desired features are for analysis purposes helping practitioners
identify and evaluate the best (sustainable) design solutions from very early stages of design.
For example, the ability to assess a building with natural ventilation was stated to be an
important design option that needs to be incorporated in to any performance sketch tool. That
the COMFEN performance sketch tool was not able to model this design feature made this a
highlighted desirable feature in the feedback.

Building visualisation in the early design interface is significant according to the use-case
feedback. It is important to practitioners – whether architects or engineers – that the interface
provides a realistic representation of the sketch design scenarios. The representation is
important for analysis and for client presentation purposes.

Cost analysis permits comparison of energy cost in use with construction costs. They were
desirable because they can help the design team to persuade the client and sell the proposed
idea (i.e. secure a design contract).

The practitioner feedback from the use-cases suggests that they want a more complex model
when using a performance sketch tool. A more complex model (e.g. imported Revit/AutoCAD
geometry or modelling of multi-zones), will take longer for the simulation engine to calculate
façade designs. Potentially, it creates a difficulty in learning/understanding how the software
works. Ultimately this complexity challenges the fundamental basis of the performance sketch
concept.

Despite this, and in contradiction to the performance sketch introduction to the use-case
sessions, the majority of the practitioners suggested that this extra precision (e.g. ability to
sketch desired geometries) in the modelling interface would be really relevant from an early
design perspective.

70
In summary, the outcome of this analysis determined three design features that require
improvement in a performance sketch tool by comparison with the COMFEN example. A
diagram of decision-making for performance sketch analysis was also established. The
following sections make up the resulting performance sketch tool specifications:

1. Freedom of Aesthetics – ability to sketch (basic) desired building geometry

2. Sustainable Designs – relevant/plausible early design performance predictions

3. Delivering the Design – realistic interface visualisation for client presentations

4. Diagram of Decision-making – guideline for performance sketch analysis

Practitioners often begin their design on a sketch pad because free-hand sketching allows them
to fully express their inspiration, aesthetically. Various designs can be sketched quickly on
paper and then presented to the client. COMFEN has the capability of comparing multiple
façades simultaneously, so was presented in the use-cases as very much a ‘sketch pad’ of
early design simulations. The presentation showed that a tool like COMFEN could be really
useful because practitioners can readily transfer their sketches into the software. Also,
COMFEN delivers the building design in a manner that is easy for the client to understand. Its
performance reporting has the potential to impress the client by quickly and graphically
presenting the costs and benefits of different design concepts.

71
4.2 Consistency Test of a Performance Sketch Tool

The consistency of design advice/feedback between performance sketch tool and detailed
design tool was evaluated by comparing the energy performance ranking of the fenestration
scenarios specified in the methodology chapter.

The following points summarise the process of energy performance ranking:

• The ranking of façades was determined by calculating the annual energy performance
in COMFEN using its one zone model and in EnergyPlus using the DoE benchmark
small and large office models (US DoE 2010) for four radically different climate zones.
• In COMFEN: the façade design with the lowest energy consumption (average
2 st
kWh/m .yr across all orientations) was ranked 1 and the façade design that
2
consumed the most energy per year (average kWh/m .yr across all orientations)
th
was ranked 4 .
• In EnergyPlus: the building design with the lowest energy consumption
2 st
(kWh/m .yr) was ranked 1 and the building that consumed the most energy per
th
year was ranked 4 .
• Plotting these façade designs left to right from first to fourth ensures that
consistency of design predictions equals ‘letter order’ in the graphs: if the ranking
in the COMFEN & EnergyPlus results has the same order then the predictions are
consistent.

4.2.1 Exploration of Possible Climatic Differences

The following table illustrates the consistency of COMFEN fenestration ranking across all four
climate zones. The table represents an overall comparison of ranking trends from country to
country; the consistency of COMFEN’s sketch advice/feedback on fenestration designs are
shown as C = Consistent Ranking (green) and I = Inconsistent Ranking (red).

Location/Climate Zone COMFEN vs. Small Office COMFEN vs. Large Office

Wellington (mixed, marine) C I

Melbourne (warm, marine) C C

Minneapolis (cold, humid) I C

Bombay (mixed, dry) C C

Table 8 – Consistency of COMFEN fenestration ranking

The results in Table 8 show that across all climate zones examined in this research COMFEN
produced consistent fenestration performance advice/feedback as multi-zone offices in
Melbourne and Bombay. However, inconsistent design messages occurred in some scenarios
in Wellington and Minneapolis.

72
The results in Figures 35~38 show the energy performance of fenestration designs in four
radically different climate zones. These locations were selected to represent a range of climate
zones around the world. They explore the differences in energy performance during sketch and
detailed design simulations.

The following are the COMFEN fenestration design scenarios which link to graphs and tables in
this section:
A. Double clear Low-E glazing with exterior venetian blinds (permanently fixed) at 45°
B. Double clear Low-E glazing with between venetian blinds (permanently fixed) at 45°
C. Double clear Low-E glazing with interior venetian blinds (permanently fixed) at 45°
D. Single clear glazing

The results in Figures 35 & 36 show the two locations where COMFEN’s calculation of the
energy performance of fenestration designs is consistent (i.e. the same letter order) with that of
EnergyPlus. However, these graphs compare the breakdown of the energy consumption from
COMFEN with the similar breakdown in EnergyPlus (EP), not just the overall total energy use.
These show that COMFEN has a very different pattern of energy use than the small office as
cooling is more important than heating. Whilst more consistent with the large office trends,
COMFEN still appears to exaggerate the importance of cooling in that the differences in cooling
energy use and ventilation between fenestration designs is far greater as a proportion of the
total energy use.

This has serious implications for the consistency of design advice. If, for example, the designer
decides to focus on what appear to be the important lessons from the COMFEN indication of
which are the most important end-uses of energy they may place too much emphasis on the
items (like ventilation in Figure 35) where the design choices appear to make a large difference
in overall energy use.

73
 300

250

200
kWh/m²/yr

Interior Equipment
150 Interior Lighting
Ventilation
100
Cooling
Heating
50

0
A B C D A B C D A B C D

COMFEN EP Small Office EP Large Office

Figure 35 – COMFEN and EP simulations in Melbourne

Figure 35 shows that the COMFEN energy performance ranking for Melbourne is consistent
with all office scenarios. Table (APPENDIX K.1) contains the results of the annual energy
distribution of designs simulated in Melbourne.

300

250

200
kWh/m²/yr

Interior Equipment
150 Interior Lighting
Ventilation
100
Cooling
Heating
50

0
A B C D A B C D A B C D

COMFEN EP Small Office EP Large Office

Figure 36 – COMFEN and EP simulations in Bombay

The Bombay results in Figure 36 show that COMFEN overall ranking is consistent with all
results derived from EnergyPlus. Table (APPENDIX M.1) contains the results of the annual
energy distribution of designs simulated in Bombay. The general lessons of the individual
breakdowns of the end-uses of energy are mostly consistent, except for case D (Single clear
glazing) where COMFEN indicates the Interior Lighting is significantly lower than EnergyPlus
suggests.

74
The results in Figures 37 & 38 show the locations where COMFEN’s ranking of energy
performance is inconsistent (i.e. a different letter order). It is evident here also that the single-
zone approach has exaggerated the distribution of energy end-uses in some scenarios.

300

250

200
kWh/m²/yr

Interior Equipment
150 Interior Lighting
Ventilation
100
Cooling
Heating
50

0
A B C D A B C D A B D C

COMFEN EP Small Office EP Large Office

Figure 37 – COMFEN and EP simulations in Wellington

Wellington COMFEN ranking (Figure 37) is consistent with majority of EnergyPlus scenarios.
However, the letter order for the large office scenario does not match COMFEN’s letter order;
Façades C & D were inconsistent (Double clear Low-E with interior venetian blinds & Single
Clear). Table (APPENDIX J.1) contains the results of the annual energy distribution of designs
simulated in Wellington. It is noticeable again that despite the consistency of the ranking of the
total energy use predictions for the different scenarios, the Heating Energy Use for the Small
Office is a significant design issue, while it is not for COMFEN or for the Large Office.

300

250

200
kWh/m²/yr

Interior Equipment
150 Interior Lighting
Ventilation
100
Cooling
Heating
50

0
A B C D C A B D A B C D

COMFEN EP Small Office EP Large Office

Figure 38 – COMFEN and EP simulations in Minneapolis

75
The Minneapolis results in Figure 38 show that the COMFEN rankings were consistent only with
the large office EnergyPlus results. The letter order was inconsistent with the small office
scenario; only fenestration D was consistent throughout all office scenarios. Table (APPENDIX
L.1) contains the results of annual energy distribution of designs simulated in Minneapolis.
Again, the underlying pattern of the energy end-uses was hugely different for the COMFEN and
Small Office cases. Even where the results and much of the breakdown patterns were similar
between COMFEN and the Large Office case, the relative significance of Heating Energy Use
was radically inconsistent (see fenestration type D).

There is an apparent trend in terms of the locations that produced consistent fenestration
ranking. Melbourne and Bombay are both hot climate zones that require a high cooling load.
Locations where inconsistent ranking occurred had cooler climatic characteristic which means
higher heating requirements: Wellington and Minneapolis.

Large amounts of energy can be saved by using appropriate fenestration design within the
building’s context. The consistency test results show that façade designs perform differently in
each location. For example, small office results in Minneapolis show some signs of
inconsistency (i.e. different ranking to COMFEN). Façade option C appears to be the most
energy efficient, which is different to other climates where façade option A is ranked first. Slight
inconsistency occurred in Wellington where option D was ranked in front of option C in the large
office scenario; COMFEN did not produce the same ranking as the large office.

It is important to note that fenestration ranking and overall energy use are only parts of any
early design advice/feedback. Distribution of energy end-uses is also highly significant because
it can influence the designer’s decisions about how to further improve the building energy
performance. Across all climates, COMFEN has the tendency to exaggerate different energy
end-uses compared to the multi-zone scenarios. This could potentially mislead designers during
early stages of design.

The scale of the building may play a large role in the determination of the distribution of energy
end-uses. For example, the distribution of floor area (i.e. percentage of perimeter & core zones)
could be a factor influencing the occurrence of inconsistencies. COMFEN does not calculate the
core zone’s energy use because it fully focuses on fenestration design with a single perimeter
zone.

76
The following table mines further into the data shown in Figure 36 for Bombay. The ranking
produced by COMFEN for Bombay (an example of a warm climate) was consistent with the
detailed model calculations:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D A B C D A B C D
Heating 0.2 0.3 0.4 0.7 0 0 0 1 0 0 0 0
Cooling 48 61 72 97 48 60 72 93 43 47 51 60
Ventilation 12 19 24 42 13 20 25 39 10 13 14 20
Interior Lighting 44 44 43 21 45 45 44 36 48 47 47 43
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 137 156 171 193 136 155 172 200 130 136 142 152
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 14% 10% 13% - 13% 11% 16% - 5% 4% 8%

Table 9 – Annual energy distribution of designs simulated in Bombay

The data bars in Table 9 show the hierarchy of energy end-uses within each fenestration design
(widest bar being the dominant consumption of energy or vice versa). Also, the table portrays
the percentage difference between fenestration designs. For example, 14% represents that the
nd
lower ranked design (e.g. 2 ) uses 14% more energy compared to the next higher ranked
st
design (e.g. 1 ). APPENDIX M contains COMFEN/EnergyPlus data and notes relating to
Bombay simulations.

Across all the climates studied, COMFEN in Bombay produced ranking of design scenario
performance predictions that were the most consistent with the multi-zone results. However, the
individual energy loads in Table 9 are exaggerated in some scenarios. It shows that COMFEN
(single-zone) exaggerates energy use in some small office scenarios:

• interior lighting load for option D is much lower (i.e. probably due to the absence of a
core zone)

The following table illustrates the inconsistent fenestration ranking produced by COMFEN in
Wellington (an example of a cool climate); COMFEN’s different distribution of energy end-uses
is evident:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D A B C D A B D C
Heating 0.9 1.0 0.6 9.4 17 15 11 19 1 1 2 1
Cooling 16 25 35 36 2 6 14 20 21 24 27 29
Ventilation 9 16 22 39 5 11 17 31 8 10 18 13
Interior Lighting 45 45 44 23 46 46 45 37 48 48 43 48
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 103 119 134 139 100 109 118 138 108 113 119 120
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 15% 13% 4% - 8% 9% 16% - 5% 6% 1%

Table 10 – Annual energy distribution of designs simulated in Wellington

The data bars in Table 10 also show the hierarchy of energy end-uses within each fenestration
design. APPENDIX J contains data and notes relating to Wellington simulations.

77
At first glance, COMFEN’s overall performance ranking is consistent with the small office
scenario. It leads to the simple conclusion that perhaps it is the floor distribution of the small
office (29% core & 79% perimeter) which is much closer to COMFEN’s perimeter zone
approach. However, the ranking order does not signify that COMFEN produces consistent
advice/feedback across the categories of energy end-use (e.g. heating & cooling requirements).
The design advice drawn from the patterns of energy end-use is shown for all the climates to be
most different for the Small Office scenario. It seems that if the single zone simulation approach
of a performance sketch tool like COMFEN is to be used, it should normally be used to model
Large Office types of building.

Early design lessons are critical. From fenestration analyses to sizing of mechanical systems,
predictions of early design performance must be reliable; the wrong design message from day
one could jeopardise the entire building project.

The simplest performance sketch approach of the single zone is apparently not adequate for
early design analyses of a multi-zone design. When examining each end-use, it is clear that the
1-zone model exaggerates some energy end-uses compared to others in the multi-zone
approach. This is undesirable, because design advice/feedback needs to be consistent in all
stages of the design process in order to help designers to identify/evaluate the best building
solutions during early stages of design.

Single zone sketch simulations require careful scrutiny to avoid the pitfalls indicated by some of
the inconsistencies that occurred in this test. Energy performance predictions derived from
COMFEN have more often than not generated fenestration rankings that are consistent across
different scenarios.

In summary, the results indicate that one zone and one façade is not enough in terms of
calculating plausible energy distribution for the whole building. The designer needs to focus on
the right design aspects from day one (e.g. cooling requirement), therefore an imprecise
distribution of the load predictions derived from a single-zone calculation is a huge issue; this
raises the question of whether it is possible to use the 1-zone approach to sketch/evaluate a
whole building, accurately.

78
4.3 Evaluating the Adequacy of the Sketch within a Performance
Sketch Tool

This section summarises the results of the analysis of the complexity of the COMFEN model as
an example of a performance sketch tool. The analysis examines how useful more complex
models might be compared to the single zone model could be in early design analyses. Options
of multi-zone simulation and natural ventilation were assessed because they were the most
commonly requested model enhancements mentioned in the conclusions of use-case studies
(also see Table 7 on page 69).

The current COMFEN interface cannot simulate multi-zone or natural ventilation models. HVAC
Only & Mixed Mode models were simulated in this study because the results illustrate whether a
multi-zone model is different to a single-zone model (i.e. does a single-zone model produce the
same design lessons/feedback with these modes incorporated?).

4.3.1 Adding Complexity to the Performance Sketch Tool

The HVAC Only models are fully conditioned by mechanical systems without any use of natural
ventilation.

The results in Figure 39 & Table 11 show the energy distribution of the ‘HVAC Only’ simulations
in Wellington. The table also illustrates the percentage differences between each modelling
method (i.e. difference between HVAC, lighting & equipment loads).

140

120

100
Interior Equipment
kWh/m²/yr

80
Interior Lighting

60 Ventilation
Cooling
40
Heating

20

0
COMFEN 2-zone 3-zone

Figure 39 – HVAC Only with daylight control simulations

79
 % Difference (+/-) % Difference (+/-)
HVAC Only 1-zone 2-zone 3-zone
1-zone vs. 2-zone 1-zone vs. 3-zone
Heating 15.5 9.8 8.6 +58% +80%
Cooling 24.8 25.8 29.4 -4% -19%
Ventilation 24.1 24.4 25.8 -1% -7%
Interior Lighting 21.8 23.6 33.3 -9% -53%
Interior Equipment 32.1 32.1 32.1 0% 0%
2
Total (kWh/m .yr) 118 116 129 +2% -9%

Table 11 – Annual energy distribution of HVAC Only models simulated in Wellington

Table 11 shows that the EUI difference between 1-zone and 2-zone models is +2%. And the
energy difference between 1-zone and 3-zone models is -9% (i.e. the COMFEN single-zone
model consumed 9% less energy than the 3-zone model).

COMFEN’s (quick) single-zone simulation method is capable of predicting overall energy

performance that is similar to the 2-zone EnergyPlus model. However, the energy differences
between 1-zone and 3-zone models are considerably larger. This is the effect of the core zone
associated to the 3-zone model, which is a zone that does not have daylight control.

The results illustrate that single-zone and multi-zone models are capable of producing similar
overall energy predictions. However, there are lessons derived from single-zone calculations
that have been exaggerated (i.e. over/below 10%) compared to multi-zone results; these are:
• 58% more heating than 2-zone; 80% more heating than 3-zone
• 19% less cooling than 3-zone
• 53% less interior lighting than 3-zone (due to no core zone in 1-zone model)

Other energy end-use differences were less than 10%.

The 1-zone model energy use was the total energy use of a north and a south facing façade
(i.e. 2 single zones), simulated separately in COMFEN. The 2-zone and 3-zone models had
lower heating load resulting from the modelling of the two zones together.

The 1-zone model consumed 53% less lighting energy than the 3-zone model because of the
daylight control. The 3-zone model contains a core zone which means daylight control is used
only in the perimeter zones. Electric lights are always used in the core zone during working
hours. Hence, the lighting load in the 3-zone model is much higher compared to the 1-zone and
2-zone models.

The Interior equipment loads across all approaches of simulation are the same because they
2
were based on a fixed simulation input (i.e. 8.1 W/m ).

80
If design advice/feedback is to be consistent, then it requires the distribution of energy end-uses
to be similar between single-zone and multi-zone models (Table 11). The results suggest that
multiple zones are required for early design analyses because the single-zone approach has
the tendency to overemphasise some energy end-uses. Potentially, the designer could be
misled (e.g. high heating requirement suggested by COMFEN).

The results in Figure 40 and Table 12 show the energy distribution of the ‘Mixed Mode’
simulations in Wellington. Mixed Mode models use natural ventilation when indoor temperature
exceeds 25°C and switched off when interior temperature drops below 18°C (i.e. HVAC is
switched off when natural ventilation is in use).

140

120

100
Interior Equipment
kWh/m²/yr

80
Interior Lighting

60 Ventilation
Cooling
40 Heating

20

0
1-zone 2-zone 3-zone

Figure 40 – Mixed Mode with daylight control simulations

% Difference (+/-) % Difference (+/-)

Mixed Mode 1-zone 2-zone 3-zone
1-zone vs. 2-zone 1-zone vs. 3-zone
Heating 15.9 12.4 9.9 +28% +61%
Cooling 23.8 23.9 28.7 0% -20%
Ventilation 24.3 24.4 26.0 0% -7%
Interior Lighting 21.8 23.6 33.3 -9% -53%
Interior Equipment 32.1 32.1 32.1 0% 0%
2
Total (kWh/m .yr) 118 116 130 +1% -10%

Table 12 – Annual energy distribution of Mixed Mode models simulated in Wellington

Table 12 shows that there is a +1% EUI difference between 1-zone and 2-zone models. And the
energy difference between 1-zone and 3-zone models is -10% (i.e. 1-zone model consumed
10% less energy compared to the 3-zone model). Again, there was virtually no difference
between 1-zone and 2-zone models. And, the energy differences between 1-zone and 3-zone
models are still considerably larger.

81
The results illustrate that the distribution of energy end-uses is considerably different between
single-zone and multi-zone models (Table 12). The Mixed Mode results also suggest that the
single-zone approach has the tendency to overemphasise some energy end-uses.

Just how close is close enough to determine that COMFEN energy predictions are plausible for
early design analyses? Clevenger & Haymaker (2006) stated:
“Energy modelling accuracy ranges from +/- 10%~40% for non-residential models
and general industry consensus is that comparisons of predicted performances are
more useful than the absolute values themselves.”

Based on this source, the single-zone prediction looks plausible as a total EUI because the
2
estimation (kWh/m .yr) was within the suggested threshold. However, some end-use differences
were above or below (+/-) the maximum value of 40% (e.g. heating & lighting energy
requirements).

Overall, the results derived from both modes illustrated that the single-zone approach has the
tendency to distort distribution of energy end-uses compared to multi-zone sketches.
Inconsistent distribution of the load predictions shows that the single-zone calculation method is
an inaccurate approach to forecast the energy distribution of a multi-zone model.

82
4.3.2 Individual Zone Comparison: Results of Adding Complexity to the Performance
Sketch Tool

This section examines the distribution of energy use within the simulation output of the ‘HVAC
Only’ models examined in the previous section. They were selected for further analysis to
illustrate the amount of energy consumed by each zone in relation to its façade orientation.

This time the energy use was broken into five major components: Heating, Cooling, Ventilation,
Interior Lighting, and Interior Equipment. The purpose was to see the energy performance of
each individual zone.

The following tables illustrate the energy consumption of each zone within all modelling
categories simulated in Wellington:
Total EUI
HVAC Only North Zone South Zone 2 EUI %
(kWh/m .yr)
Heating 4.2 11.3 15.5 13%
Cooling 21.0 3.8 24.8 21%
Ventilation 18.6 5.5 24.1 20%
Interior Lighting 11.6 10.1 21.8 18%
Interior Equipment 16.0 16.0 32.1 27%
2
Total EUI (kWh/m .yr) 71.5 46.7 118 100%
EUI % 61% 39% 100% -

Table 13 – Distribution of energy use for each zone within the COMFEN model

Total EUI
HVAC Only North Zone South Zone 2 EUI %
(kWh/m .yr)
Heating 2.6 7.1 9.7 8%
Cooling 22.0 3.9 25.9 22%
Ventilation 18.8 5.6 24.4 21%
Interior Lighting 13.4 10.2 23.6 20%
Interior Equipment 16.0 16.0 32.1 28%
2
Total EUI (kWh/m .yr) 72.9 42.8 116 100%
EUI % 63% 37% 100% -

Table 14 – Distribution of energy use for each zone within the 2-zone model

Total EUI
HVAC Only North Zone Core Zone South Zone 2
EUI %
(kWh/m .yr)
Heating 2.9 0.5 5.2 8.6 7%
Cooling 19.3 7.3 2.8 29.4 23%
Ventilation 18.1 5.3 2.5 25.9 20%
Interior Lighting 6.6 16.7 10.0 33.2 26%
Interior Equipment 10.7 10.7 10.7 32.1 25%
2
Total EUI (kWh/m .yr) 57.5 40.5 31.2 129 100%
EUI % 45% 31% 24% 100% -

Table 15 – Distribution of energy use for each zone within the 3-zone model

83
 2
The total energy use (kWh/m .yr) is the sum of all the zones in a model. The EUI percentage
column in the tables indicates what proportion each end-use is of the total EUI. The EUI
percentage row at the bottom of each table shows the fraction of the total EUI contributed by
each zone. There are some large differences in the EUI percentage columns between the 1-, 2-
and 3-zone models, communicating in particular very different messages about the significance
of heating energy use.

Results in Table 13 show that COMFEN’s north zone utilizes 61% of the total EUI. And, the
south zone consumes approximately 39% of the total annual energy.

Table 14 illustrates that the 2-zone model’s north zone utilises 63% of the total EUI; and the
south zone consumes approximately 37% of the total annual energy.

Table 15 shows that the 3-zone model’s north zone utilizes 45% of the total EUI. The south
zone consumes 24% of the total energy and the core zone at approximately 31%.

It is apparent that the core zone has a significant effect on the distribution of energy end-uses.
The lesson learned from the modelling categories is that a multi-zone model (i.e. zones
simulated together) behaves differently to a model with zones simulated separately. This
suggests that the single-zone approach is missing the connection between thermal zones; it is
simply not enough to sum up the total of ‘two’ single zones.

The major difference can be seen in the heating, cooling and ventilation loads. In general,
perimeter zones require more energy to condition because they are closer to the exterior
elements compared to the core zone space. The heat gains and losses are greater, particularly
through the fenestration. The north zone consumes more cooling and ventilation energy
2
(kWh/m .yr) because in Wellington the north facing zone is exposed to direct sunlight more
throughout the year compared to the south facing zone. By contrast, the core zone relies on
electric lighting the entire day. This is because there is minimal or no natural light entering the
space.

The following table summarises the EUI percentage of energy distribution within each modelling
category in Tables 13~15.

HVAC Only 1-zone 2-zone 3-zone

Heating 13% (16) 8% (10) 7% (9)
Cooling 21% (25) 22% (26) 23% (29)
Ventilation 20% (24) 21% (24) 20% (26)
Interior Lighting 18% (22) 20% (24) 26% (33)
Interior Equipment 27% (32) 28% (32) 25% (32)
2
Total EUI % & (Total EUI kWh/m .yr) 100% (118) 100% (116) 100% (129)

Table 16 – EUI percentage comparison of the modelling categories simulated in Wellington

84
Table 16 illustrates that the COMFEN single zone model can produce similar total EUI as multi-
zone models. Ventilation and interior equipment loads are similar across all modelling
categories. However, some design lessons derived from COMFEN are not the same in the
multi-zone models. These are:
• The 1-zone model has a much higher heating load than 2-zone & 3-zone models
• The cooling load is considerably lower than the 3-zone approach
• The interior lighting load is much lower than that shown by the 3-zone approach:
the core zone is critical.

Results in Table 16 illustrate the benefit of daylighting in 1-zone and 2-zone models. These
models therefore had a lower percentage of lighting energy use compared to the 3-zone model.

In general, the EUI of all modelling categories were similar. However, it is clear that the
percentage of energy distribution changes considerably in multi-zone simulations; especially
when a core zone is modelled (i.e. heating & lighting requirements).

Overestimated/underestimated predictions could potentially mislead the designer from day one.
The option to simulate multiple zones needs to be incorporated into COMFEN and other similar
tools in order performance sketch predictions to be more relevant during the early design stage.

85
4.3.3 A Consistency Test for the Study of Adding Complexity to the Performance
Sketch Tool

The analysis examines single-zone and multi-zone models with natural ventilation incorporated.
The results illustrate the consistency of sketch design advice/feedback between single-zone
and multi-zone models.

The following are the fenestration designs examined in this test (see Figure 34):
A. Double clear Low-E glazing with exterior venetian blinds at 45° (permanently fixed)
B. Double clear Low-E glazing with between venetian blinds at 45° (permanently fixed)

2 st
The façade design with the lowest EUI (kWh/m .yr) is ranked 1 . Consistency is achieved if
single-zone and multi-zone predictions of energy performance generate the same design
advice/feedback (i.e. the façade ranking and the distribution of energy end-uses is similar).

120

100

80
kWh/m²/yr

Interior Equipment
60 Interior Lighting
Ventilation
Cooling
40
Heating

20

0
A B A B A B

1-zone 2-zone 3-zone

Figure 41 – Consistency test: Mixed Mode models with daylight control

86
The following table illustrates façade ranking and energy consumption (EUI percentages) of the
fenestration designs tested in the Wellington:

Simulation Category 1-zone 2-zone 3-zone

Façade Code A B A B A B
2.7 2.2 1.7 1.4 1.0 1.0
Heating
(3%) (2%) (2%) (1%) (1%) (1%)
11 18 12 19 15 21
Cooling
(11%) (16%) (12%) (17%) (14%) (19%)
9 14 10 14 10 14
Ventilation
(9%) (12%) (9%) (12%) (9%) (12%)
44 44 45 45 44 44
Interior Lighting
(45%) (40%) (45%) (40%) (44%) (39%)
32 32 32 32 32 32
Interior Equipment
(32%) (29%) (32%) (29%) (32%) (29%)
2 99 109 101 111 102 112
Total EUI (kWh/m .yr) & EUI%
(100%) (100%) (100%) (100%) (100%) (100%)
Ranking 1 2 1 2 1 2
% Difference - 11% - 10% - 10%

Table 17 – Consistency test: Mixed Mode models with daylight control

The results in Figure 41 & Table 17 show that the façade rankings (letter orders) are consistent
st
throughout the categories of modelling method (i.e. ranked 1 = Double clear Low-E with
nd
exterior venetian blinds & ranked 2 = Double clear Low-E with between venetian blinds).

Results in Table 17 illustrates that the 1-zone Performance Sketch can produce fenestration
ideas consistent with those from multi-zone models. However, some design lessons derived
from the 1-zone model are different. These are:
• much higher 1-zone heating load than 2-zone & 3-zone models
• 1-zone cooling load lower than the 3-zone model

Ventilation, interior equipment and lighting loads are similar across the modelling categories.

In this test the Interior Lighting percentages were really similar across all simulation categories;
fenestration designs had a huge influence on the distribution of energy end-uses (i.e. effect of
permanently fixed venetian blinds).

Overall, results illustrate that a one zone model can produce a similar Total EUI as multi-zone
sketches. However, similar to other performance sketch tests in this research the multi-zone
models produced a different design message when the focus is on the distribution of load
predictions. The differences might be small in this test, but as the scale of a model increases,
there is a possibility that the differences could be immense; this suggests the importance of
modelling multiple zones during performance sketch analyses.

87
88
5 Practitioner-based Performance Sketch Tool Specifications

The use-cases suggested a user desire for increased modelling complexity in future versions of
a performance sketch tool. The study of the influence of the complexity of the performance
sketch tool model, suggested there is a need for that complexity if the energy performance is to
relate to full scale. At their simplest, actions resulting from these recommended actions would
make more technological functions of EnergyPlus available in COMFEN simulation. If the detail
of the sketch tool is increased users will be able to simulate more complex façade designs and
derive extra relevant predictions to utilise in the detailed design.

The Performance sketch concept requires a simplified model of a building - a (performance)

sketch. However, the exploration of the single-zone COMFEN model as an example of a
performance sketch tool suggests the single-zone approach is inadequate. Possibly, it can be a
misleading sketch of the thermal performance of a whole building.

The use-cases suggest that despite understanding the performance sketch concept the
practitioners still desired a more detailed sketch. This raises a serious question as to whether
any simple approach to evaluating the essential features of a building could be used in
association with a detailed thermal model.

The outcome of the use-case analyses determined three areas of design feature in COMFEN
that users suggested required improvement in future versions. As a summary of this, a diagram
of decision-making for performance sketch analysis was also established. The final sections of
this chapter describe in turn each of these individual areas within the performance sketch tool
specifications:

1. Freedom of Aesthetics – ability to sketch (basic) desired building geometry

2. Sustainable Designs – relevant/plausible early design messages/predictions

3. Delivering the Design – good interface graphical visualisation for client presentations

4. Diagram of Decision-making – guideline for performance sketch analysis

The specifications are based on the desired COMFEN features suggested in the use-case
studies. These specifications can help the development of future versions of performance
sketch tools; and assist users during the early stages of design.

89
5.1 Freedom of Aesthetics

The use-cases suggest that it is vital for users of the performance sketch tool to have the ability
to express (sketch) their desired building geometry. The following specifications are related to
freedom of aesthetics in early design tools.

Ability to sketch any desired building geometry:

• The designer should be able to express their ideas without restrictions during the
sketch design phase.

Visualisation of the sketched design within the interface:

• The sketched building geometry should be realistic, visual elements such as glass
colour, solar shading devices, and surface materiality are important to the user.

Ability to apply/select construction materials:

• A performance sketch tool should allow the user to pick construction material for
each building surface in the sketched geometry (i.e. walls/ceiling/floor/roof).

Capability of importing complex architectural geometry:

• The performance sketch tool should allow the user to sketch complex shapes in
software such as Revit and ArchiCAD/AutoCAD, and then import the file into
COMFEN for performance analysis.

Ability to export COMFEN models:

• The performance sketch tool should allow the user to export the sketched design
to other common detailed design tools (e.g. EnergyPlus – compatible file format).

90
5.2 Sustainable Designs

Practitioners suggested that it is vital for an early design tool to have the capability of examining
a wider range of design scenarios than COMFEN can currently analyse. This they recognised is
at variance with the fundamentals of the performance sketch concept. Ultimately, this new
approach to modelling would add huge complexity to a performance sketch tool.

The following specifications are related to analysis options required by COMFEN users in early
design tools:

Cost analysis:

• A performance sketch tool should allow the user to plot in cost of energy (e.g.
$/kWh)

Design options:

• Natural ventilation, double façade and atrium options should be available in a tool
for more sustainable building designs.

Shading devices:

• Freedom of geometry in designing of custom shading devices.

External shading:

• A performance sketch tool should allow the user to simulate the proposed design
in context, with neighbouring buildings and trees etc.

HVAC systems:

• Option to select the type of HVAC system and (at least) set seasonal efficiency for
summer & winter.

Multi-zone simulations:

• A performance sketch tool should allow the user to understand the effects of air
and heat movement between perimeter and core zones in the building design.
And, it should have the ability to simulate floor sections within a building design
(i.e. top, middle & bottom floors).
• In Figure 42, illustrates the option of a ‘Generic Core Zone’, where the user can
2
plot in mechanical load values per metre square (W/m ) for the electricity
consumption (i.e. HVAC, lighting & interior equipment etc). Essentially, with the
additional core zone plus four perimeter zones, energy performance of a whole
(multi-zone) floor/building could be examined in COMFEN (i.e. stated in the use-
cases).

91
Figure 42 – COMFEN perimeter zones with an additional generic core zone

CZ = Core Zone & PZ = Perimeter Zone

Modelling environment:

• A performance sketch tool should allow the user to examine the proposed design
in context with surrounding buildings and trees (present) in the simulation model;
this will truly show the effects of the surrounding elements on the proposed
design’s performance.

5.3 Delivering the Design

The use-case results suggest that communication with the client is paramount in order for the
project to progress smoothly. The following specifications are related to visualisation (for the
client in particular) of predictions within early design tools.

Visualisation of design:

• A performance sketch tool should allow the user to explore the building design
(i.e. orbit around the 3D model).
• A performance sketch tool should allow the user to clearly present the proposed
design to the client, aesthetically.
• The proposed design should be supported by easy to understand graphical
information relating to overall performance (i.e. tables & figures).

92
5.4 Diagram of Decision-making for Performance Sketch Analysis

The following is an expanded version of Kalay’s diagram (Figure 15) of workflow and subtasks
in performance-based design. The original diagram has been modified to incorporate
performance sketch tools. It should be noted that this diagram identifies two parts of the overall
design process where the performance sketch tool might be applied: i) conventionally – during
sketch design; and ii) during detailed design when some design option that could be sketched
can be studied rapidly through the sketch.

Figure 43 – Diagram of decision-making for performance sketch analysis

This diagram reflects the way in which architects and engineers perform performance sketch
analysis. It is intended that it guide tool developers in production of new (performance sketch)
analytical tools. It provides a framework in which detailed design tools like Energy Plus and
OpenStudio can be planned and developed in a parallel manner.

93
94
6 CONCLUSIONS

The following returns to the research questions stated in the aim & objectives section of
Chapter 1:

1. What are practitioners’ needs from a performance sketch tool?

Specifications in Chapter 5 illustrate the ideal model of a performance sketch early

design tool.

2. How relevant is the information predicted by a performance sketch tool?

The use-cases suggest that COMFEN predictions are valuable. However, the

capability of only sketching orthogonal-shaped designs and not being able to model

the gamut of sustainable design ideas (e.g. atrium design & natural ventilation etc) is

a big restriction for practitioners (i.e. less relevant to complex designs).

3. Performance Sketch Early Design Tool – how simple is too simple?

Currently, based on the use-cases single-zone is said to be efficient (quick) and

easy to understand. However, many have suggested that there is a need to increase

the complexity in geometry modelling (i.e. freedom to sketch anything).

4. How close is close enough?

The single-zone sketch has the tendency to over emphasise (distort) distribution of

energy load predictions when compared to multi-zone results derived from

EnergyPlus.

The use-case studies in this research established the definition of an ideal Performance Sketch
early design (simulation) tool. Despite understanding the intention of the performance sketch
concept, architects and engineers still desire that extra precision in the COMFEN modelling
interface. The use-cases suggested that in order the sketch design advice/feedback to be more
relevant, the designer must have the freedom to sketch/evaluate design elements such as
building geometry, HVAC system, building schedules, lighting loads and occupancy.

COMFEN and EnergyPlus fenestration design ranking comparison determined the consistency
between sketch predictions and detailed calculations. In summary, the tests in this research
indicate that it is inadequate to solely focus on the fenestration design when examining the
energy performance of a whole building, accurately (i.e. multiple zones have to be assessed
during the early design stage). The calculation method of one zone and one façade has
demonstrated that it cannot produce plausible distribution of energy load predictions; this
imprecise design message could mislead the designer from day one.

95
The Performance Sketch Specifications can help improve future simulation tools’ usability and
capability by applying suggestions stated in the interview use-cases; also assist designers
performing sketch analyses.

Practitioners rely on early design tools to forecast the energy performance of a proposed
design. It is crucial to get the basic fundamentals right from day one of the project, because
problems that occur later on could be costly to the client and design team.

COMFEN could evolve (develop) into a software that sketches ‘multi-zone buildings’ from the
early design stage. Also, the possibility of having natural ventilation calculations in COMFEN, by
having (blank) boxes for users to enter (the desired) input values; basic (natural ventilation)
equation that links with the modelled geometry. Therefore, architects and engineers can take
these sketch (multi-zone) predictions into the detailed design stage for further development.

In conclusion, it is clear that the performance sketch tool requires further improvements
(i.e. plausible distribution of energy end-uses). And experience is required in order to examine
real building designs with COMFEN and EnergyPlus. However, COMFEN can certainly develop
into a powerful performance sketch tool in the future because of its compatibility with the
EnergyPlus simulation engine.

96
FUTURE WORK

The outcome of this thesis suggests that the following research questions require further
testing and investigation:

• Test a wide range of climate zones within COMFEN to see whether inconsistent
design advice/feedback only occurs in the climate zones demonstrated in this
research; with detailed examination of the distribution of energy end-uses.

• Design features such as double façade and natural ventilation being incorporated
into upcoming versions of COMFEN.

• Development of a performance sketch early design tool that is capable of testing

‘Non-linear’ shaped building designs (i.e. increase complexity in geometry
modelling).

COMFEN has the potential to be a power simulation tool for early design analyses; in the
course of further development, COMFEN could be a very efficient way to sketch/evaluate the
energy performance of numerous commercial buildings in New Zealand.

BRANZ’s Building Energy End-use Study (BEES) will produce reliable data on building
schedules, material properties, equipment loads, and occupancy. For the first time anywhere it
will be possible to simulate/test a building based upon the median/mean, standard deviation; or
some percentile expression of the real measured data of New Zealand commercial buildings.

The BEES data is important for the Performance Sketch because it offers the opportunity to
critique the performance sketch concept with real buildings/data; this will help improve
software usability and capability.

97
98
REFERENCES

American Society of Heating, R. a.-c. (2005). 2005 ASHRAE handbook : fundamentals. Atlanta:
American Society of Heating, Refridgeration and Air-conditioning Engineers (ASHRAE).

Attia, S., Beltrán, L., De Herde, A., & Hensen, J. (2009). Architect Friendly: a comparison of ten
different building performance simulation tools. Eleventh International IBPSA Conference (pp.
204-211). Glasgow: Building Simulation.

Augenbroe, G. (2002). Trends in building simulation. Building & Environment , 891-902.

Bambardekar, S. (2009). The architect as performer of energy simulation in the early design
stage. Eleventh International IBPSA Conference (pp. 1306-1313). Glasgow: Building
Simulation.

Bambardekar, S., & Poerschke, U. (2009). The architect as performer of energy simulation in
the early design stage. Eleventh International IBPSA Conference (pp. 1306-1313). Glasgow:
Building Simulation.

Bannister, P. (2009). Why good buildings go bad while some are just born that way. Exergy
Australia Pty Ltd.

Bordass, B., & Leaman, A. (1997). From Feedback to Strategy. Retrieved 2011, from
http://www.usablebuildings.co.uk:
http://www.usablebuildings.co.uk/Probe/ProbePDFs/PCW5.pdf

Brahme, R., O’Neill, Z., Sisson, W., & Otto, K. (2009). Using existing whole building energy tools
for designing Net-Zero energy buildings - challenges and workarounds. Eleventh International
IBPSA Conference (pp. 9-16). Glasgow: Building Simulation.

Brock, L. (1948). Designing the exterior wall: an architectural guide to the vertical envelope .
Hoboken, New Jersey : John Wiley & Sons, Inc.

Clevenger, C. M., & Haymaker, J. (2006). The impact of building occupant on energy modelling
simulations. Joint International Conference on Computing and Decision Making in Civil and
Building Engineering. Montreal.

Commission of the European Communities. (1993). Use of sophisticated building energy

simulation tools. 3rd European Conference on Architecture "Solar energy in architecture and
urban planning" . Florence.

Crosbie, M. J. (2005). Curtain walls: recent developments by Cesar Pelli & Associates. Basel:
Birkhauser.

Deplazes, A., Elsener, C., Roesler, S., Seher, C., & Seigrist, T. (2008). Constructing
Architecture - Materials, Processes, Structures (Second Edition). Berlin: Birkhauser.

99
Eisele, J., & Kloft, E. (2003). High-Rise Manual: Typology and Design, Construction and
Technology . Basel: Birkhauser.

Fabrizio, E., Corrado, V., & Filippi, M. (2010). A model to design and optimize multi-energy
systems in buildings at the design concept stage. Renewable Energy , 644–655.

Fernandez, N. P. (2008). The influence of construction materials on life-cycle energy use and
carbon dioxide emissions of medium size commercial buildings. Wellington.

Feustel, H. E. (1989). The Comis infiltration model, proceedings building simulation. Vancouver:
IBPSA.

Flanagan, R., & Norman, G. (1983). Life cycle costing for construction. London: Surveyors
Publications.

Gan, G. H. (2000). Effective depth of fresh air distribution in rooms with single-sided natural
ventilation. Energy and Buildings , 65-73.

Goulding, J. e. (1993). Energy in architecture: The Passive Solar Handbook. Dublin:

Commission of the European Communities.

Gratia, E., & De Herde, A. (2002). A simple design tool for the thermal study of an office
building. Energy and Buildings , 279-289.

Hari, A. (2001). Integrated, Customer Driven, Conceptual Design Methodology, Doctoral Thesis.
Haifa: Israel Institute of Technology.

Haumer, P. (2004). Use Case-Based Software Development. In I. Alexander, & N. Maiden,

Scenarios, Stories, Use Cases: Through the Systems Development Life-Cycle.

Hausladen, G., de Saldanha, M., & Liedl, P. (2006). ClimateSkin: building skin concepts that
can do more with less energy . Berlin.

Hensen, J. (1994). Energy related design decisions deserve simulation approach. Proc.
International Conference on Design and Decision Support Systems in Architecture & Urban
Planning. Eindhoven.

Hensen, J., Djunaedy, E., Radošević, M., & Azzedine, Y. (2004). Building performance
simulation for better design: some issues and solutions . The 21th Conference on Passive and
Low Energy Architecture, (pp. 1-6). Eindhoven.

Hindrichs, D. U., & Heusler, W. (2010). Façades - Building Envelopes for the 21st Century: 3rd
expanded edition. Berlin: Birkhauser.

Hitchcock, R. J., Mitchell, R., Yazdanian, M., Lee, E., & Huizenga, C. (2008). COMFEN: a
commercial fenestration façade design tool. Third National Conference of IBPSA-USA (pp. 246-
252). Berkeley: SimBuild.

100
Hoes, P., Hensen, J. L., Loomans, M., de Vries, B., & Bourgeois, D. (2009). User behavior in
whole building simulation. Energy and Buildings , pp. 295–302.

Hong, & Tianzhen. (1997). Building Simulation: Frequently.

Hsu, C.-Y., & Donn, M. (2009). Commercial Building Façade Design: The relationship between
early design lessons and detailed design lessons”. ANZAScA. Launceston.

Hui, S. C. (1998). Simulation based design tools for energy efficient buildings in Hong Kong.
Hong Kong Papers in Design and Development .

Hui, S., & Cheung, K. (1998). Application of building energy simulation to air-conditioning
design. Mainland-Hong Kong HVAC Seminar, (pp. 12-20). Beijing.

Hyde, R., & Pedrini, A. (1996). An energy conservation architectural design tool for warm
climate (LTV): the tool development and testing.

Judkoff, R., & Neymark, J. (2006). Model validation and testing: The methodological foundation
of ASHRAE Standard 140 (2006). ASHRAE Transactions, 112 PART 2 , 367-376.

Kalay, Y. E. (1999). Performance-based design, Automation in Construction. 395-409.

Kolokotroni, M., Robinson-Gayle, S., Tanno, S., & Cripps, A. (2004, January-February).
Environmental impact analysis for typical office façades. Building Research and Information ,
pp. 3, 15.

Lain, M., Hensen, J., & Zmrhal, V. (2009). Computer simulation for better design and operation
of large office building air-conditioning. Eleventh International IBPSA Conference (pp. 740-745).
Glasgow: Building Simulation.

Lam, K., Huang, Y., & Zhai, C. (2004). Energy Modeling Tools Assessment For Early Design
Phase. Pittsburgh.

LBNL. (2010, April). COMFEN 3.0. Retrieved April 2010, from LBNL Window and Daylighting
Software: http://windows.lbl.gov/software/comfen/3/index.html

Leaman, A., & Bordass, B. (1999). Productivity in buildings: the `killer’ variables. Building
Research & Information .

Lewis, J. R. (1994). Sample sizes for usability studies: Additional considerations. Human
Factors , 368-378.

Lewis, M. (2004). Integrated design for sustainable buildings. ASHRAE.

Lomanowski, B. A., & Wright, J. L. (2007). Modelling fenestration with shading devices in
building energy simulation: a practical approach. Eleventh International IBPSA Conference (pp.
976-983). Glasgow: Building Simulation.

Lstiburek, J. W. (2008). Why green can be wash. ASHRAE Journal , 28-36.

101
Mazzarella, L., & Pasini, M. (2009). Building energy simulation and object-oriented modelling:
review and reflections upon achieved results and further developments. Eleventh International
IBPSA Conference (pp. 638-645). Glasgow: Building Simulation.

Nielsen, J. (1993). Usability Engineering. London: Academic Press.

Nielsen, J., & Landauer, T. K. (1993). A mathematical model of the finding of usability problems.
ACM INTERCHI’93 Conference (pp. 206-213). Amsterdam: ACM Press.

Ochoa, C. E., & Capeluto, I. G. (2008). Advice tool for early design stages of intelligent façades
based on energy and visual comfort approach. Energy and Buildings , pp. 480–488.

Papakonstantinou, K. A., Kiranoudis, C. T., & Markatos, N. C. (2000). Numerical simulation of

air flow field in single-sided ventilated buildings. Energy and Buildings , 41-48.

Papamichael, K., & Ellington, K. (1994). Computer-Based Design Tools: Reaching

decisionmakers during the building design process.

Persily, A. K. (1993). Envelope Design Guidlines for Federal Office Buildings: Thermal Integrity
and Airtightness. pp. 2.1-1, 2.2-2, 2.2-3, 2.3-5, 2.4-4.

(2010). Personal communication from my thesis supervisor: video recording of Performance

Sketch Tool presentation. (M. Donn, Interviewer)

Petersen, S., & Svendsen, S. (2010). Method and simulation program informed decisions in the
early stages of building design. Energy and Buildings .

Pollock, M., Roderick, Y., McEwan, D., & Wheatley, C. (2009). Building simulation as an
assisting tool in designing an energy efficient building: a case study. Building Simulation, (pp. 1-
12). Glasgow.

Prasad, B. N., & Bhat, A. K. (2005). Computer Aided Building Energy Simulation. IE(I) Journal--
AR , 28-31.

Punjabi, S., & Miranda, V. (2005). Development of an integrated building design information
interface. IBPSA. Montreal.

Rowe, G., & Wright, G. (1999). The Delphi technique as a forecasting tool: Issues and analysis.
International Journal of Forecasting , 353-375.

Shaviv, E., Yezioro, A., Capeluto, I. G., Peleg, U. J., & Kalay, Y. E. (1996). Simulations and
knowledge-based computer-aided architectural design (CAAD) systems for passive and low
energy architecture. Energy and Buildings , pp. 257-269.

Skulmoski, G. J., Hartman, F. T., & Krahn, J. (2007). The Delphi Method for Graduate
Research. Journal of Information Technology Education , 1-21.

102
Tianzhen, H., Jinqian, Z., & Yi, J. (1997). IISABRE: An Integrated Building Simulation
Environment. Building and Environment , 219-224.

Torcellini, P., Deru, M., Griffith, B., & Benne, K. (2008). DOE Commercial Building Benchmark
Models. California: U.S. Department of Energy.

Turner, C. W., Lewis, J. R., & Nielsen, J. (2006). Determining Usability Test Sample Size.
Human Factors , 3084-3088.

Tzempelikos, A., & Athienitis, A. K. (2005). Integrated Thermal and Daylighting Analysis for
Design of Office Buildings. ASHRAE Transactions: Research , p. 227.

Tzempelikos, A., Athienitis, A. K., & Karava, P. (2007). Simulation of façade and envelope
design options for a new institutional building. Solar Energy , pp. 1088–1103.

US Department of Energy. (2010). EnergyPlus Energy Simulation Software. Retrieved February

10, 2010, from US Department of Energy - Energy Efficeincy and Renewable Energy:
http://apps1.eere.energy.gov/buildings/energyplus/

US Department of Energy. (2007). Technical report.

Virzi, R. A. (1992). Refining the test phase of usability evaluation: How many subjects is
enough? Human Factors , 457-468.

Voss, K., Herkel, S., Pfafferott, J., Lohnert, G., & Wagner, A. (2007). Energy efficient office
buildings with passive cooling – Results and experiences from a research and demonstration
programme. Solar Energy , 424–434.

Walton, G. N. (1989). Airflow network models for element-based building airflow modelling.
ASHRAE.

Witzig, A., Foradini, F., Probst, M., & Roecker, C. (2009). Simulation tool for architects:
optimization of active and passive solar use. International conference CISBAT, (pp. 1-7).
Lausanne.

Wong, N. H., Mahdavi, A., Boonyakiat, J., & Lam, K. P. (2003). Detailed multi-zone air flow
analysis in the early building design phase. Building and Environment , 1-10.

Wood, A. (2010). Best Tall Buildings 2009 - CTBUH International Award Winning Projects. New
York: Routledge.

Yohanis, Y. G., & Norton, B. (2002). The early design model for prediction of energy and cost
performance of building design options. International Journal of Sustainable Energy , 47-61.

103
104
APPENDICES

APPENDIX A – Selection of Practitioners: Information Sheet

Project Title – Commercial Building Façade Design: Improving the Consistency of Early Design
Tool Predictions and Detailed Design Tool Calculations

To whom it may concern,

I am a student from The School of Architecture at Victoria University of Wellington, planning to work with
practitioners to determine their simulation needs during the early stages of design. These are use-cases,
which are interview feedback from users of COMFEN (building energy simulation program). The use-cases
will play a huge role in terms of improvements to early design tools like COMFEN. The feedback is based
on the introduction of COMFEN to practitioners, so they will determine their needs based on past/current
project experiences by stating the essential information they need for the early and detailed stages of
design. These improvements will make the early design tool more user-friendly. The aim of this research
project is to improve the consistency between early design tools and detailed design tools and to develop
a list of façade design specifications (template models) for early design tools that are based on feedback
from interviews with practitioners (use-cases).

Improvements in early design tool simulations has the potential to save money and time during the early
design stage. The capability to simulate multiple façade designs quickly and accurately will benefit clients
and practitioners in the New Zealand building industry. It will provide architects and engineers with trust,
confidence and reliable design messages that can be carried out into the detailed design stage.

Your approval is required to undertake this research project. I am interviewing practitioners like yourself
from various firms. You will answer a series of questions. Notes will be taken during the interviews and
also a voice recording. The introduction of COMFEN and interview will take approximately 40 minutes
(single visit).

Should any participants feel the need to withdraw from the project, they may do so without question at any
time before the data is analysed. Email or call me if that’s the case.

The gathered information and voice recordings will remain confidential and deleted at the completion of the
report (it will not be possible for you to be identified personally in the report). No other person besides
myself and my supervisor, Michael Donn, will see the raw interview feedback.

A copy of the final report can be provided to you, if you wish.

Yours sincerely,
Henry Hsu

Enquiries:
Chi‐Yao (Henry) Hsu Michael Donn (research supervisor)
Contact number: +64 21 1013243 Contact number: +64 21 611 280
Email: [email protected] Email: [email protected]

105
APPENDIX B – Selection of Practitioners: Consent Form

Victoria University of Wellington

Consent to Participation in Research
Project Title – Commercial Building Façade Design: improving the relevance of early design
tools to detailed design tools

Please tick:

I agree to take part in this research

I consent to information or opinions which I have given being quoted anonymously in

any reports on this research

I would like the tape recordings of my interview returned to me at the conclusion of the
project

I understand that I will have an opportunity to check the transcripts of the interview
before publication

I understand that the data I provide will not be used for any other purpose or released to
others without my written consent

I would like to receive a summary of the results of this research when it is completed

To receive a copy of the report

Name of participant (please print clearly)............................................................

Signature...................................................... Date......................................................

106
APPENDIX C – Practitioner A: Initial Interview Feedback
th
9 July 2010

COMFEN Advantages:

• COMFEN Narrow down façade options quickly, so practitioners can move on to the next
design phase.

Desired features in future versions of COMFEN:

• The design options costed

• Double skin façade option needed

• Need for natural ventilation simulations

• Adjust natural ventilation openings

• External shading, neighbouring buildings and trees etc needed

• Ability to have angled fins and shades

• The option to plot in values

•
2
Able to change pump and fan values, W/m etc

• Fan energy, it’s detailed for a simple tool

• Visual drawing of the fins and shading devices, glazing on the section and plan views
etc, realistic image to convince the people view the results and design

• Visual feedback, shading device needs to be visible on the interface

• Peak heating and cooling for summer and winter (seasonal efficiency)

Comments which challenge the idea of a Performance sketch tool:

• HVAC, what type is it in detailed, more description and info

• Simulate what you want, reduce limitation of COMFEN

• Variables in the HVAC system (various systems to choose from)

107
APPENDIX C.1 – Practitioner A: Analysis of Initial Interview
Feedback

Occupation: Architect & Engineer

Summary of COMFEN Advantages:

• COMFEN narrow down façade options quickly, so practitioners can move on to the next
design phase.

Summary of Conclusions:

• Desire of cost analysis in the interface.

• Desire of natural ventilation, double façade and atrium design option.

• Desire of more complex HVAC systems available in COMFEN. Currently, the

program lacks precision from the HVAC system perspective.

• Option of simulating external shading, neighbouring buildings and trees.

• Realistic image of shading devices and fins shown on the sketched scenarios for
aesthetics visualisation.

• Simulate what you want in terms of geometry, reduce limitations in COMFEN.

Analysis of Conclusions:

• A complex (HVAC) model will take more time for the simulation engine to
calculate façade designs. Also, it will create difficulty on learning/understanding
how the software works (i.e. time consuming), which ultimately challenges the
performance sketch idea. However, the question is (if possible): just how helpful
would this extra precision in the modelling procedure be to the early design
stage?

• For example, if a curved façade is graphically represented differently on the

interface (i.e. with multiple straight lines), is that a big problem in terms of the
aesthetics? Or as a practitioner, it is absolutely vital for the digital model to look
realistic? Regardless if the simplified shape represents the equivalent idea of the
actual design.

108
APPENDIX C.2 – Practitioner A: 2nd Round of Feedback
I definitely think that if it is to be a successful tool for early façade analysis there shouldn't be
much limitation on façade options in which to compare. So all types of glazing, external shading,
double façade options should be available. Agree that HVAC system type isn't as important for
just façade performance analysis.

I don't think that cost analysis of options needs to be included. My revised comments are as
follows:

Design conditions

• Need to have control over design room heating and cooling temperatures and not rely
on COMFEN defaults

• Need to have control over design ambient temperatures for peak heating and cooling
load calc

• Should be able to use NIWA weather files for NZ site locations for energy calc

Glazing

• Allow simple definition as Window to Wall Ratio (WWR), G-value, Overall U-value
(including frame effect)

Wall

• Allow option to define as simple R-value

External shading

At a minimum external shading should capture the following:

• Fixed vertical elements

• Fixed horizontal elements

• Movable shading - open or closed based on incident solar radiation

• Should be able to see in 3D model and orbit around

It also would be good to be able to "draw in" shading elements rather than just define
dimensions.

Internal blinds

• Define by shading coefficient

• Control operation based on incident solar radiation

Internal gains

• Need to have control over inputs and operational profiles and not just rely on COMFEN
defaults

109
HVAC

Don’t think the actual HVAC system type is required to compare façade options but in order
to benchmark annual energy use it is important to be able to input both:
• Heating system seasonal efficiency e.g. boiler = 0.95, heat pump = 3.0

• Cooling system seasonal efficiency e.g. chiller = 5.0

Otherwise if they are both assumed to have an efficiency of 1.0 it can change the comparative
performance between options and resulting hierarchy of façade performance.

Would then be good to have the option to model an HVAC system in slightly more detail by
including:
•
2
Outdoor air rate lit/s per m

•
2
Annual pump and fan energy kWh/m

• Heat recovery effectiveness % (may require additional inputs to define)

• Economiser cycle (may require additional inputs to define)

That is a simple way to capture different HVAC systems and in terms of calculation shouldn’t be
too complex

Ventilation

• As above don’t think necessary to include outdoor air rate for just façade performance
comparisons

•
2
Need to include natural ventilation option - define as free net open-able area (m ),
operation profile based on internal temperature and wind speed at a minimum

• Mixed mode ventilation option should also be available

If I were to design a tool it would also have these additional features:

• Produce easy to read printout of all model inputs (or notional inputs selected by the
user)

• Calculate frequency of thermal comfort range (PMV) during occupied hours

• Automatically compare to NZBC (NZS4243) reference model performance for site

location

•
2
Convert peak heating and cooling loads to W/m format and show on the same chart

•
2
Automatically combine annual heating and cooling energy and convert to kWh/m per
year format

• Automatically produce charts comparing multiple façade options. Also allow output in
table format to bring into MS Excel

• Automatically include façade summary description of each option on chart e.g. WWR,
glazing U-value, glazing G-value, shading description

• Automatically include a small image of each façade option on charts as well

110
APPENDIX C.3 – Practitioner A: Analysis of 2nd Round of Feedback
Judging by the feedback of Practitioner A, it is very clear that the performance sketch tool needs
to increase its complexity. However, compromises can be made through incorporating the
nd
essentials stated in the 2 round of feedback without losing the sketch characteristic of
COMFEN (i.e. features noted in the summary of conclusions).

st
The question asked after the 1 round of feedback was: just how helpful would this extra
precision in the (HVAC) modelling process be to the early design stage? The question was
asked because it challenged the idea of performance sketch. Practitioner A has pointed out the
importance of the option for HVAC seasonal efficiency, because without it (i.e. current version of
COMFEN) the default settings could cause hierarchy in the façade performance predictions.
Therefore, this feature is essential for future versions of COMFEN (i.e. more input options for
HVAC modelling). In this case, longer calculation time with increase of HVAC complexity is
preferred over the speed provided by general default settings. Hindsight, it might take the user
longer to learn/understand how the software works.

Another question asked after the initial feedback was: it is absolutely vital for the model to look
nd
realistic? Practitioner A emphasised in the 2 round of feedback that various design options
should be available in the software (e.g. double façade & natural ventilation). Visualisation of
the designs in the interface is vital. The user should not be restricted to only modelling
orthogonal shaped designs. Essentially, allow the user to freely sketch the desired building
geometry or the equivalent shape. Restrictions of sketching should be avoided during design
stage because it will result in aesthetics limitation.

In summary, Practitioner A agreed (settled) with most of the conclusions stated in the initial
feedback in relation to HVAC complexity and freedom of sketching; whereas, the costing option
was later said to be unnecessary. The feedback pointed out the significance of the need to
simulate the desired building shape and its mechanical system (i.e. more relevant predictions).
Also, the relationship between visual presentation and simulation results is paramount in order
the user to understand the predicted data.

111
APPENDIX D – Practitioner B: Initial Interview Feedback

rd
3 August 2010

COMFEN Advantages:

• It is very good to have the ability to compare multiple façade designs at once that are
supported by graphical results.

Desired features in future versions of COMFEN:

• Costing and material analysis

• Natural ventilation needed

• Shading device modelling (its effects). Modelling of complex shapes (curved), analysis
of its transmission level and solar gain, thermal mass

• Double façade and atrium design option

• Lux level on the ceiling, light bouncing off the shading device, opportunity to use the
combination of electric lighting and natural lighting

• colour of glass shown on the sketched scenario

Comments which challenge the idea of a Performance sketch tool:

• Option to alter the interior wall/ceiling/floor material and colour

• Ability to import Revit (or other drawing software) complex geometries into COMFEN (at
the moment it lacks of precision from the modelling perspective)

• Limited in terms of shape capable of modelling

• Need to analyse the air movement, heat movement between perimeter zones and core
zone, because it may be a different result once the core zone is incorporated in detailed
design, which is too late. Therefore need multi-zone simulation during early design
stage.

• It’s just a comparative tool at the moment, will move on quickly from here, we need to
know what the full building would behave like earlier on

112
APPENDIX D.1 – Practitioner B: Analysis of Initial Interview
Feedback

Occupation: Architect

Summary of COMFEN Advantages:

• It is very good to have the ability to compare multiple façade designs at once that are
supported by graphical results.

Summary of Conclusions:

• Desire of cost analysis in the interface.

• Desire of natural ventilation, double façade and atrium design option.

• Desire of importing complex architectural geometry into COMFEN (e.g. curved

shading devices drawn in Revit or other drawing software). COMFEN currently
lacks precision from the modelling perspective.

• Colour of glass should be shown on the sketched scenario for aesthetics

visualisation.

• The need to have multi-zone simulations because the air/heat movement between
the perimeter zones and core zones could influence the design predictions
dramatically. Therefore we need to know how the ‘full building’ would behave like.

Analysis of Conclusions:

• A complex (Revit) model will take more time for the simulation engine to calculate
façade designs. Also, it will create difficulty on learning/understanding how the
software works, which ultimately challenges the performance sketch idea.
However, just how helpful would this extra precision in geometry modelling be to
the early design stage?

• So if a curved shading device is graphically represented differently on the

interface (i.e. with straight lines instead), is that a big problem in terms of
visualisation? Is it enough to have simplified shapes that represent the equivalent
design idea? Or as practitioner, it is absolutely vital for the shading device to look
realistic within a digital model?

No further feedback was given by Practitioner B, therefore initial feedback remains as their final
statement on COMFEN as a performance sketch tool.
113
APPENDIX E – Practitioner C: Initial Interview Feedback

th
25 June 2010

COMFEN Advantages:

• It is great that COMFEN (early design tool) is compatible with EnergyPlus.

• Also, the plugin with Google SketchUp makes sketch design user-friendly. The
COMFEN program seems user-friendly which is a plus.
• The information generated by COMFEN is enough and useful from an early design
perspective. CO2 analysis very useful during early design stage.
• COMFEN and EnergyPlus can be used at the same time to determine the
desired/optimum glazing type and building geometry.

Desired features in future versions of COMFEN:

• Double façade and atrium design option

• Adding costs to the program is a good idea

Comments which challenge the idea of a Performance sketch tool:

• to have full building simulation during early design stage (multi-zone simulation in
COMFEN not available).

• be more complex, desire of doing simple “full” building simulations in COMFEN.

Because COMFEN at the moment seems just like a program that determines the
façade design (that’s what it is for), however, it is not all about glazing/façade type
because the core zone(s) of the building design will affect the overall energy
consumption, as well as atrium spaces.

114
APPENDIX E.1 – Practitioner C: Analysis of Initial Interview
Feedback

Occupation: Architect & Engineer

Summary of COMFEN Advantages:

• Compatible with EnergyPlus.

• Opentstudio Plug-in with Google SketchUp makes sketch design user-friendly.
• COMFEN predictions is enough and useful from an early design perspective.
• CO2 analysis very useful.
• COMFEN and EnergyPlus can be used at the same time to determine the
desired/optimum glazing type and building geometry.

Summary of Conclusions:

• Desire of cost analysis in the interface.

• Desire of double façade and atrium design option.

• Desire of doing simple “full” building simulations in COMFEN.

Analysis of Conclusions:

• A complex (multi-zone) model will take more time for the simulation engine to
calculate façade designs. Also, it will create difficulty on learning/understanding
how the software works, which ultimately challenges the performance sketch idea.
However, just how helpful would this extra precision in ‘building zoning’ be to the
early design analysis?

No further feedback was given by Practitioner C, therefore initial feedback remains as their final
statement on COMFEN as a performance sketch tool.

115
APPENDIX F – Practitioner D: Initial Interview Feedback

th
6 September 2010

COMFEN Advantages:

• Speed of simulation is fast, quick analysis

• ASHRAE PoE comfort calculation good for Green Star PoE assessment. Also, the
daylight tab is good information. Good for early design stage decision making.
• COMFEN provides enough info from an early design stage perspective. Good balance.

Desired features in future versions of COMFEN:

• Need natural ventilation option (multi-zone when have NV), define openable windows.
Dynamic model, mechanical windows that adjust according to the temperature and
weather.

• HVAC type, its parameters, more info on exactly what the HVAC package is.

• Ability to define set points temperature etc winter summer

• Need more geometry freedom for aesthetics purposes

• Import geometry from different drawing software

• Export COMFEN model into common detailed design tool that are used by firms in the
world, file format that is compatible.

• Architects like to have a realistic image of the design – freedom of sketching and
displayed on the interface. Sections and elevations. Ability to alter the façade, angles
and aesthetic pleasing shapes.

Comments which challenge the idea of a Performance sketch tool:

N/A

116
APPENDIX F.1 – Practitioner D: Analysis of Initial Interview
Feedback

Occupation: Engineer

Summary of COMFEN Advantages:

• Speed of simulation
• ASHRAE Post Occupancy Evaluation (POE) comfort calculation is good for Green
Star POE assessments.
• Daylight predictions are good for early design stage decision making.
• COMFEN provides enough information from an early design stage perspective
(i.e. good balance).

Summary of Conclusions:

• Need natural ventilation option (multi-zone when have NV), define open-able
windows. Dynamic model, mechanical windows that adjust according to the
temperature and weather.

• HVAC type, its parameters, more info on exactly what the HVAC package is.

• Ability to define set points temperature etc winter summer

• Need more geometry freedom for aesthetics purposes

• Import geometry from different drawing software

• Export COMFEN model into common DDT that are used by firms in the world, file
format that is compatible.

• Architects like to have a realistic image of the design – freedom of sketching and
displayed on the interface. Sections and elevations. Ability to alter the façade,
angles and aesthetic pleasing shapes.

Analysis of Conclusions:

• Practitioner D stated that COMFEN is a good simulation tool overall for the sketch
design phase. However, there were several design options that need to be
incorporated into future versions of COMFEN in order to meet the desires (seen in
summary of conclusions). Freedom of sketch and realistic visualisation are vital
design aspects for Practitioner D; this suggests there is a need for improvements
in these areas.

Practitioner D had no further feedback, therefore initial feedback remains as their final
statement on COMFEN as a performance sketch tool.

117
APPENDIX G – Practitioner E: Initial Interview Feedback

nd
2 June 2010

COMFEN Advantages:

If design is more complex then COMFEN is good tool for sketch design, non symmetrical
windows etc.

COMFEN is a good tool for communicating with the client, seal the deal with client with its
graphical interface.

If cost too high, but still needs to meet the client’s requirement, therefore minor changes will be
done on the façade design, this is where COMFEN can come in handy/useful.

More useful/suitable for material/window manufacture

Desired features in future versions of COMFEN:

• Double façade and atrium design option

• Need Life Cycle Cost calculation in COMFEN (more detail), way to convince client (the
proposed design) with the interface, LCC graphs, design benefits etc.

• Cost estimation, annual cost estimation, benefits to client, payback cost/time

• Façade material (need to have the option to change - the opaque areas)

• Cost of materials etc, thermal performance, its life expectancy (blank box to fill in value)

Comments which challenge the idea of a Performance sketch tool:

• It’s useful (COMFEN) for architects but it’s not really part of my job. I will recommend
various designs, but the performance analysis is for engineers/manufactures.

118
APPENDIX G.1 – Practitioner E: Analysis of Initial Interview
Feedback

Occupation: Architect

Summary of COMFEN Advantages:

• A good tool for sketch design (e.g. non-symmetrical windows on building façade).
• COMFEN is a good for communicating with the client, with its graphical interface.
• If cost too high, but still needs to meet the client’s requirement, this is where
COMFEN can come in handy/useful.
• More useful/suitable for window manufacture

Summary of Conclusions:

• Desire of Life Cycle Cost analysis in the interface.

• Desire of double façade and atrium design option.

• COMFEN is useful for architects, but it’s not really part of my job.

Analysis of Conclusions:

• Practitioner E suggests that COMFEN is a useful tool for what it is, but its current
functionality is more relevant to window manufactures rather than architects.
• For example, if it was possible to simulate ‘anything you want’ but the graphics of
these design choices are not displayed exactly in COMFEN (i.e. simplified
geometry of the actual design), is that a big problem in terms of the aesthetics?
Or as practitioners, it is absolutely vital for the model to look realistic? Regardless
the simplified shape which represents the equivalent idea of the actual design.
Also, would it be more ‘part of your job’ if the program expands its design
capability?

No further feedback was given by Practitioner E, therefore initial feedback remains as their final
statement on COMFEN as a performance sketch tool.

119
APPENDIX H – Practitioner F: Initial Interview Feedback

16/09/2008 (recorded USA interview)

COMFEN Advantages:

Good comparison options, ability to compare four designs/scenarios at once (neat).

Interface – how façade designs link to graphs and simulation output data: very important
for architects, understanding the data and good graphic presentation.

Desired features in future versions of COMFEN:

• Current design opinions – limitation of aesthetics? (Modelling parameters)

• Aesthetic pleasing & sustainable designs that look good and perform well
• Sell the idea to the client by indicating the benefits of their investment, short term & long
term (design punch line).
• Glass fins, rotating shading devices, double façade option
• Solar gain and heat gain analyses
• Selection of glass colour (aesthetic visualisation of design)
• Selection of building standards (design requirements)
• Shading and sun path study

Comments which challenge the idea of a Performance sketch tool:

• Why is one zone enough?

120
APPENDIX H.1 – Practitioner F: Analysis of Initial Interview
Feedback

Occupation: Architect

Summary of COMFEN Advantages:

• Ability to compare four designs/scenarios at once

• Graphical information

Summary of Conclusions:

• Ability to sketch freely

• Glass fins, rotating shading devices, double façade option
• Solar gain and heat gain analyses
• Selection of glass colour (aesthetic visualisation of design)
• Selection of building standards (design requirements)
• Shading and sun path study

Analysis of Conclusions:

• Practitioner F suggests that COMFEN is a neat tool. However, raised a question:

why is one zone enough? The 1-zone method is specifically for analyses of
façade performance; focusing on types of glazing & shading device.
• Freedom of sketch and visualisation are vital aspects for the interface; this
suggests there is a need for improvements in these areas.

121
APPENDIX I – Façade Constructions & Material Properties
Each fenestration option was constructed by a list of building materials and the chosen shading
device. The following fenestration options were used in the consistency test of COMFEN 3.0,
where the layers of construction go from outer layer to inner layer in each list:

(A) Double clear Low-E glazing with exterior venetian blinds at 45°

• Exterior shading device = horizontal metal venetian blinds, permanently fixed at 45

degrees (1mm thick)
• Outer layer = 3.175mm Low-E glazing
• Air gap = 12.7mm (argon & air)
• Inner layer = 5.715mm Low-E glazing

(B) Double clear Low-E glazing with between venetian blinds at 45°

• Outer layer = 3.175mm Low-E glazing

• Air gap = 6.4mm (argon & air)
• Between shading device = horizontal metal venetian blinds, permanently fixed at 45
degrees (1mm thick)
• Air gap = 6.4mm (argon & air)
• Inner layer = 5.715mm Low-E glazing

(C) Double clear Low-E glazing with interior venetian blinds at 45°

• Outer layer = 3.175mm Low-E glazing

• Air gap = 12.7mm (argon & air)
• Inner layer = 5.715mm Low-E glazing
• Interior shading device = horizontal metal venetian blinds, permanently fixed at 45
degrees (1mm thick)

(D) Single clear glazing

• Outer layer = 5.664mm clear glazing

• Air gap = none
• Inner layer = none
• Shading device = none

122
APPENDIX J – Simulation Location: Wellington, New Zealand
(Mixed, marine climate zone)
Façade Code: A North South East West Average
Heating 0.7 1.1 0.9 0.9 0.9
Cooling 17 15 16 16 16
Ventilation 9 8 8 9 9
Interior Lighting 44 48 45 45 45
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 102 104 103 104 103

J. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Wellington

Façade Code: B North South East West Average

Heating 0.6 1.4 0.9 0.9 1.0
Cooling 28 16 26 28 25
Ventilation 17 9 17 23 16
Interior Lighting 43 48 45 44 45
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 120 106 121 128 119

J. 2 – COMFEN Double clear Low-E between VB 45 simulated in Wellington

Façade Code: C North South East West Average

Heating 0.4 0.8 0.6 0.7 0.6
Cooling 40 22 37 41 35
Ventilation 22 10 24 33 22
Interior Lighting 42 47 44 43 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 137 112 138 150 134

J. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Wellington

Façade Code: D North South East West Average

Heating 8 15 8 8 9.4
Cooling 49 10 40 45 36
Ventilation 40 11 43 61 39
Interior Lighting 22 24 23 23 23
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 150 92 147 169 139

J. 4 – COMFEN Single clear simulated in Wellington

123
APPENDIX J.1 – Simulation Location: Wellington, New Zealand
The following table shows the energy performance ranking and energy distribution of each
fenestration design in the Wellington climate zone:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D A B C D A B D C
Heating 0.9 1.0 0.6 9.4 17 15 11 19 1 1 2 1
Cooling 16 25 35 36 2 6 14 20 21 24 27 29
Ventilation 9 16 22 39 5 11 17 31 8 10 18 13
Interior Lighting 45 45 44 23 46 46 45 37 48 48 43 48
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 103 119 134 139 100 109 118 138 108 113 119 120
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 15% 13% 4% - 8% 9% 16% - 5% 6% 1%

J. 1. 1 – Annual energy distribution of designs simulated in Wellington

The percentage difference illustrates the amount of energy consumed more by the lower ranked
nd
fenestration design. In other words, 15% represents that the lower ranked design (e.g. 2 ) uses
st
15% more energy (annually) compared to the higher ranked design (e.g. 1 ).

st nd
The COMFEN results show that there is a 15% difference between the 1 (A) and 2 (B)
fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-E
nd rd
with between venetian blinds). The difference between the 2 (B) and 3 (C) options is 13%
(Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior venetian
rd th
blinds). The difference between the 3 (C) and 4 (D) options is very close at 4% (Double clear
Low-E with interior venetian blinds vs. Single Clear).

st
The EnergyPus results of the small office show that there is an 8% difference between the 1
nd
(A) and 2 (B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double
nd rd
clear Low-E with between venetian blinds). The difference between the 2 (B) and 3 (C)
options is 9% (Double clear Low-E with between venetian blinds vs. Double clear Low-E with
rd th
interior venetian blinds). The difference between the 3 (C) and 4 (D) options is 16% (Double
clear Low-E with interior venetian blinds vs. Single Clear).

st
The EnergyPus results of the large office show that there is a 5% difference between the 1 (A)
nd
and 2 (B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double
nd rd
clear Low-E with between venetian blinds). There is a 6% difference between the 2 (B) and 3
(D) options (Double clear Low-E with between venetian blinds vs. Single clear). The difference
rd th
between the 3 (D) and 4 (C) options is 1% (Single clear vs. Double clear Low-E with interior
venetian blinds).

COMFEN predictions were consistent across all scenarios in the small office category; the
performance advice/feedback was the same. Slight inconsistency occurred when option D was
ranked in front of option C in the large office category; COMFEN did not produce the same
(consistent) design advice/feedback as the large (multi-zone) office. The table illustrates that
COMFEN tends to exaggerate distribution of energy end-uses compared to some multi-zone
scenarios (e.g. higher cooling requirements).

124
The following shows the energy end-use ranking within each COMFEN fenestration design in
the Wellington climate zone:
st
1 Fenestration A: interior lighting was the dominant end-use, followed by cooling,
interior equipment, ventilation and heating.

nd
2 Fenestration B: interior lighting was the dominant end-use, followed by cooling,
ventilation, interior equipment, and heating.

rd
3 Fenestration C: interior lighting was the dominant end-use, followed by cooling,
interior equipment, ventilation, and heating.

th
4 Fenestration D: ventilation was the dominant end-use, followed by cooling,
interior equipment, interior lighting, and heating.

The following shows the energy end-use ranking within each fenestration design in the small
office category in the Wellington climate zone.
st
1 Fenestration A: interior lighting was the dominant end-use, followed by interior
equipment, heating, ventilation, and cooling.

nd
2 Fenestration B: interior lighting was the dominant end-use, followed by interior
equipment, heating, ventilation, and cooling.

rd
3 Fenestration C: interior lighting was the dominant end-use, followed by interior
equipment, ventilation, cooling, and heating.

th
4 Fenestration D: interior lighting was the dominant end-use, followed by interior
equipment/ventilation, cooling, and heating.

The following shows the energy end-use ranking of each fenestration design in the large office
category in the Wellington climate zone:
st
1 Fenestration A: interior lighting was the dominant end-use, followed by interior
equipment, cooling, ventilation, and heating.

nd
2 Fenestration B: interior lighting was the dominant end-use, followed by interior
equipment, cooling, ventilation, and heating.

rd
3 Fenestration D: interior lighting was the dominant end-use, followed by interior
equipment, cooling, ventilation, and heating.

th
4 Fenestration C: interior lighting was the dominant end-use, followed by interior
equipment/cooling, ventilation, and heating.

125
APPENDIX K – Simulation Location: Melbourne, Australia
(Warm, marine climate zone)
Façade Code: A North South East West Average
Heating 0.6 1 0.9 0.9 0.8
Cooling 19 17 18 18 18
Ventilation 10 10 10 11 10
Interior Lighting 43 47 45 45 45
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 105 107 106 107 106

K. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Melbourne

Façade Code: B North South East West Average

Heating 0.5 1.3 0.9 0.9 0.9
Cooling 30 19 28 31 27
Ventilation 18 11 19 26 18
Interior Lighting 42 47 44 44 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 122 110 124 133 122

K. 2 – COMFEN Double clear Low-E between VB 45 simulated in Melbourne

Façade Code: C North South East West Average

Heating 0.4 0.7 0.6 0.8 0.6
Cooling 42 25 40 44 37
Ventilation 23 13 27 36 25
Interior Lighting 41 47 43 43 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 139 117 143 156 138

K. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Melbourne

Façade Code: D North South East West Average

Heating 9 15 9 9 10
Cooling 54 16 47 53 42
Ventilation 43 16 49 69 44
Interior Lighting 22 23 23 22 22
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 159 103 160 185 152

K. 4 – COMFEN Single clear simulated in Melbourne

126
APPENDIX K.1 – Simulation Location: Melbourne, Australia
The following table shows the energy performance ranking and energy distribution of each
fenestration design in the Melbourne climate zone:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D A B C D A B C D
Heating 0.8 0.9 0.6 10.4 14 12 9 16 1 1 1 2
Cooling 18 27 37 42 3 9 17 26 23 25 30 30
Ventilation 10 18 25 44 7 13 20 36 9 12 14 20
Interior Lighting 45 44 44 22 46 45 45 37 48 48 47 43
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 106 122 138 152 100 110 121 145 109 115 122 125
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 15% 13% 10% - 10% 10% 20% - 5% 6% 3%

K. 1. 1 – Annual energy distribution of designs simulated in Melbourne

st nd
In the COMFEN results there is a 15% difference between the 1 (A) and 2 (B) fenestration
options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-E with between
nd rd
venetian blinds). The difference between the 2 (B) and 3 (C) options is 13% (Double clear
Low-E with between venetian blinds vs. Double clear Low-E with interior venetian blinds). The
rd th
difference between the 3 (C) and 4 (D) options is 10% (Double clear Low-E with interior
venetian blinds vs. Single Clear).

st
In EnergyPus results of the DoE small office there is a 10% difference between the 1 (A) and
nd
2 (B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear
nd rd
Low-E with between venetian blinds). The difference between the 2 (B) and 3 (C) options is
10% (Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior
rd th
venetian blinds). The difference between the 3 (C) and 4 (D) options is 20% (Double clear
Low-E with interior venetian blinds vs. Single Clear).

st nd
In EnergyPus results of the DoE large office there is a 5% difference between the 1 (A) and 2
(B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-
nd rd
E with between venetian blinds). There is a 6% difference between the 2 (B) and 3 (C)
options (Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior
rd th
venetian blinds). The difference between the 3 (C) and 4 (D) options is 3% (Double clear Low-
E with interior venetian blinds vs. Single clear).

Overall, COMFEN energy performance predictions in Melbourne were consistent with all office
scenarios. However, the table illustrates that COMFEN tends to exaggerate distribution of
energy end-uses compared some multi-zone scenarios (e.g. lower heating requirements).

127
APPENDIX L – Simulation Location: Minneapolis-St Paul, USA
(Cold, humid climate zone)
Façade Code: A North South East West Average
Heating 25 21 23 23 23
Cooling 14 15 15 15 15
Ventilation 11 11 11 13 12
Interior Lighting 48 43 45 45 45
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 130 122 127 128 127

L. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Minneapolis

Façade Code: B North South East West Average

Heating 27 15 19 20 20
Cooling 15 25 25 26 23
Ventilation 12 17 21 27 19
Interior Lighting 48 42 44 44 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 135 131 141 149 139

L. 2 – COMFEN Double clear Low-E between VB 45 simulated in Minneapolis

Façade Code: C North South East West Average

Heating 24 12 14 16 16
Cooling 20 36 34 37 32
Ventilation 14 21 27 36 25
Interior Lighting 47 41 43 43 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 137 142 152 164 149

L. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Minneapolis

Façade Code: D North South East West Average

Heating 98 52 55 54 65
Cooling 15 51 44 47 39
Ventilation 22 45 63 84 53
Interior Lighting 24 22 23 22 23
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 192 201 218 240 212

L. 4 – COMFEN Single clear simulated in Minneapolis

128
APPENDIX L.1 – Simulation Location: Minneapolis-St Paul, USA
The following table shows the energy performance ranking and energy distribution of each
fenestration design in the Minneapolis climate zone:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D C A B D A B C D
Heating 23 20 16 65 84 105 99 138 10 10 9 24
Cooling 15 23 32 39 14 4 8 23 18 21 25 25
Ventilation 12 19 25 53 28 18 23 52 9 12 14 23
Interior Lighting 45 44 44 23 45 45 45 37 48 48 48 43
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 127 139 149 213 201 204 205 280 115 119 125 145
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 10% 7% 43% - 1% 1% 37% - 4% 4% 16%

L. 1. 1 – Annual energy distribution of designs simulated in Minneapolis

st nd
In the COMFEN results there is a 10% difference between the 1 (A) and 2 (B) fenestration
options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-E with between
nd rd
venetian blinds). The difference between the 2 (B) and 3 (C) options is 7% (Double clear
Low-E with between venetian blinds vs. Double clear Low-E with interior venetian blinds). The
rd th
difference between the 3 (C) and 4 (D) options is 43% (Double clear Low-E with interior
venetian blinds vs. Single Clear).

st
In EnergyPus results of the DoE small office there is a 1% difference between the 1 (C) and
nd
2 (A) fenestration options (Double clear Low-E with interior venetian blinds vs. Double clear
nd rd
Low-E with exterior venetian blinds). The difference between the 2 (A) and 3 (B) options is
1% (Double clear Low-E with exterior venetian blinds vs. Double clear Low-E with between
rd th
venetian blinds). The difference between the 3 (B) and 4 (D) options is 37% (Double clear
Low-E with between venetian blinds vs. Single Clear).

st nd
In EnergyPus results of the DoE large office there is a 4% difference between the 1 (A) and 2
(B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-
nd rd
E with between venetian blinds). There is a 4% difference between the 2 (B) and 3 (C)
options (Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior
rd th
venetian blinds). The difference between the 3 (C) and 4 (D) options is 16% (Double clear
Low-E with interior venetian blinds vs. Single clear).

The percentage difference between fenestration designs closes drastically in the office
categories (i.e. there is almost no difference between designs A, B & C).

Overall, COMFEN energy performance predictions were consistent with the large office
EnergyPlus results. However, the COMFEN letter order was inconsistent with the small office
category; only fenestration D was consistent throughout all office scenarios. However, the table
illustrates that COMFEN tends to exaggerate distribution of energy end-uses compared to some
multi-zone office scenarios (e.g. lower heating requirements).

129
APPENDIX M – Simulation Location: Bombay, India
(Mixed, dry climate zone)
Façade Code: A North South East West Average
Heating 0.2 0.2 0.2 0.2 0.2
Cooling 47 48 48 49 48
Ventilation 11 11 12 13 12
Interior Lighting 46 43 44 44 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 137 135 136 139 137

M. 1 – COMFEN Double clear Low-E exterior VB 45 simulated in Bombay

Façade Code: B North South East West Average

Heating 0.2 0.2 0.3 0.5 0.3
Cooling 52 58 64 69 61
Ventilation 13 13 21 28 19
Interior Lighting 46 42 43 43 44
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 144 145 161 173 156

M. 2 – COMFEN Double clear Low-E between VB 45 simulated in Bombay

Façade Code: C North South East West Average

Heating 0.3 0.2 0.4 0.7 0.4
Cooling 58 67 77 85 72
Ventilation 15 15 27 38 24
Interior Lighting 46 41 43 42 43
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 152 155 180 198 171

M. 3 – COMFEN Double clear Low-E interior VB 45 simulated in Bombay

Façade Code: D North South East West Average

Heating 0.4 0.3 0.9 1.3 0.7
Cooling 68 85 112 124 97
Ventilation 22 20 53 74 42
Interior Lighting 21 20 21 21 21
Interior Equipment 32 32 32 32 32
2
Total (kWh/m .yr) 143 158 219 252 193

M. 4 – COMFEN Single clear simulated in Bombay

130
APPENDIX M.1 – Simulation Location: Bombay, India
The following table shows the energy performance ranking and energy distribution of each
fenestration design in the Bombay climate zone:

Simulation Category COMFEN EP Small Office EP Large Office

Facade Code A B C D A B C D A B C D
Heating 0.2 0.3 0.4 0.7 0 0 0 1 0 0 0 0
Cooling 48 61 72 97 48 60 72 93 43 47 51 60
Ventilation 12 19 24 42 13 20 25 39 10 13 14 20
Interior Lighting 44 44 43 21 45 45 44 36 48 47 47 43
Interior Equipment 32 32 32 32 31 31 31 31 29 29 29 29
Total (kWh/m2.yr) 137 156 171 193 136 155 172 200 130 136 142 152
Ranking 1 2 3 4 1 2 3 4 1 2 3 4
% Difference - 14% 10% 13% - 13% 11% 16% - 5% 4% 8%

M. 1. 1 – Annual energy distribution of designs simulated in Bombay

st nd
In the COMFEN results there is a 14% difference between the 1 (A) and 2 (B) fenestration
options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-E with between
nd rd
venetian blinds). The difference between the 2 (B) and 3 (C) options is 10% (Double clear
Low-E with between venetian blinds vs. Double clear Low-E with interior venetian blinds). The
rd th
difference between the 3 (C) and 4 (D) options is 13% (Double clear Low-E with interior
venetian blinds vs. Single Clear).

st
In EnergyPus results of the DoE small office there is a 13% difference between the 1 (A) and
nd
2 (B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear
nd rd
Low-E with between venetian blinds). The difference between the 2 (B) and 3 (C) options is
11% (Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior
rd th
venetian blinds). The difference between the 3 (C) and 4 (D) options is 16% (Double clear
Low-E with interior venetian blinds vs. Single Clear).

st nd
In EnergyPus results of the DoE large office there is a 5% difference between the 1 (A) and 2
(B) fenestration options (Double clear Low-E with exterior venetian blinds vs. Double clear Low-
nd rd
E with between venetian blinds). There is a 4% difference between the 2 (B) and 3 (C)
options (Double clear Low-E with between venetian blinds vs. Double clear Low-E with interior
rd th
venetian blinds). The difference between the 3 (C) and 4 (D) options is 8% (Double clear Low-
E with interior venetian blinds vs. Single clear).

Overall, COMFEN energy performance predictions were consistent with all results derived from
EnergyPlus. Bombay and Melbourne were the only locations where COMFEN predictions were
entirely consistent with the office scenarios.

However, the table illustrates that COMFEN tends to exaggerate distribution of energy end-uses
compared to some multi-zone office scenarios (e.g. higher cooling requirements).

131