MARI LIIS MARIPUU Licentiate Thesis

THESIS FOR THE DEGREE OF LICENTIATE OF ENGINEERING
Adapting Variable Air Volume (VAV) Systems for Office

Buildings without Active Control Dampers
Function and Demands for Air Distribution Components
MARI-LIIS MARIPUU
Building Services Engineering

Department of Energy and Environment
CHALMERS UNIVERSITY OF TECHNOLOGY
Göteborg, Sweden 2006
Adapting Variable Air Volume (VAV) Systems for Office Buildings
without Active Control Dampers
MARI-LIIS MARIPUU
© MARI-LIIS MARIPUU
Technical report D2006:02

Department of Energy and Environment
Chalmers University of Technology
SE-412 96 Göteborg
Sweden
Telephone +46(0) 31-772 1000
ISSN: 1652-6007
Printed by
Chalmers Reproservice
Göteborg 2006
ii
Adapting Variable Air Volume (VAV) Systems for Office
Buildings without Active Control Dampers
MARI-LIIS MARIPUU
Chalmers University of Technology
ABSTRACT
In commercial buildings the building services systems may account for a substantial
part of the need of energy. There is, at least in Sweden, a large existing stock of 20 to
30 year old office buildings, which are supplied with centralised constant air volume
flow (CAV) air conditioning systems with the primary function to keep the room
temperature at an acceptable level summertime. However, since the heat loads are only
occasionally at the peak level, considerable amounts of energy can be wasted during
periods when cooling is not needed or when rooms are not in use.
Substantial energy savings can be achieved by changing the system to variable air
volume flow (VAV) operation. Nevertheless, such a change becomes meaningful only
if the system as a whole is adapted to the new configuration. The aim of this work is to
clarify the prerequisites for a well functioning VAV system in office buildings. A
special focus has been that the system solutions should be uncomplicated and possible
to implement in existing buildings as well as in new ones.
Adopting VAV in office buildings without having to install active control dampers in
the duct system is one way of building up a simple VAV system. To obtain this, some
special demands must be set upon the VAV variable supply air diffusers. They must
have good controlling properties and a low noise level even at a high pressure drop
over the device. Also, a comfortable airflow pattern in the room must be ensured
within the variable volume flow range in question.
One example of uncomplicated system configuration, which in theory has properties

that fulfil the demands, was tested both in the laboratory and studied in plants in
function, with focus on the indoor climate and the need of energy. Furthermore the
influence of a variable air volume flow on duct heat gains was analysed and
mathematically evaluated.
The results indicate that it is clearly possible to obtain a well functioning system with
a technical solution in accordance with the studied one. The indoor climate demands
are essentially fulfilled. The noise levels in rooms can be kept at a well acceptable
level. The draught appeared to be considerably low despite low supply air
temperatures. The field monitoring of energy consumption showed that the tested
VAV systems worked energy efficiently. The heat recovery system managed to cover
all the heating need of the supply air, i.e. any additional heating with the heating coil
was not needed even at low outdoor temperatures.
Keywords: CAV, VAV, variable supply air diffuser, climate control, energy
efficiency, ventilation
iii
This report refers to the research project “CAVA” at Chalmers University of
Technology, grant no. 218 98-1, funded by Swedish Energy Agency. The report also
refers to Alfred Ots Scholarship Foundation
iv
PREFACE
This work has been carried out at the division of Building Services Engineering,
Department of Energy and Environment, Chalmers University of Technology. The
work is a part of a pilot project CAVA (From Constant Air Volume to Variable Air
Volume), which has been conducted under the EUFORI program funded by Swedish
Energy Agency (Statens Energimyndighet). The project CAVA has been carried out in
co-operation with the following companies: CIT Energy Management, Akademiska
Hus, Bravida, Lindinvent, Energo, Swedish Energy Agency.
I would like to acknowledge the generous and constructive help I have received from a
large number of people during this thesis work. I would like to thank the company
representatives of the above mentioned companies for their guidance and input to this
project. In particular, a special thanks goes to Thomas Lindborg from Lindinvent, for
his help and technical support for my laboratory and field measurements.
I would like to express my gratitude to my supervisor Prof. Emeritus Enno Abel for
his support, constructive feedback and dedication during this work. This thesis would
have been impossible to perform without him, since his help and enthusiasm made my
studies in Chalmers University of Technology possible. My gratitude goes also to
Prof. Emeritus Mart Mägi and Prof. Emeritus Sven Olving who, through the Alfred
Ots Scholarship Foundation , granted me the financial support and therewith supported
my studies and stay in Sweden.
I wish to thank my assistant supervisor, Assoc. Prof. Lennart Jagemar for his
invaluable guidance, considerable support and enormous patience with solving my
academic problems. I highly appreciate the support and valuable contribution from
Prof. Per Fahlén.
I would like to thank Leif Nilsson, Tommy Sundström and Josef Jaroz for providing
technical support and assistance in my measurements. I would also like to express my
gratitude to all of my colleagues at Building Service Engineering for creating a nice
and friendly working environment. A special thanks goes to Katarina Bergkvist for her
invaluable assistance.
Finally I would like to send warm thanks to my family and friends for their continuous
support during these years. A special thanks goes to my BEST-Gothenburg team, who
have made my stay in Sweden unforgettable.
Mari-Liis Maripuu
Gothenburg, May 2006
v
vi
TABLE OF CONTENTS
ABSTRACT……………………………………………………………………….…iii
PREFACE…………………………………………………………………………….v
TABLE OF CONTENTS…………………………………………………………...vii
NOMENCLATURE.…………………………………………………………………ix
1 INTRODUCTION……………………………………………………………..1
1.1 Background…………………………………………………………………...1
1.1.1 General…………………………………………………….................1
1.1.2 CAV and VAV……………………………………………..................2
1.1.3 The supply air temperature…………………………………............2
1.2 Objectives and scope…………………………………………….................3
1.3 Methodology…………………………………………………………………..4
1.4 Outline of the thesis……………………....................................................5
1.5 Literature review……………………………………………………………...5
2 VAV AS A SYSTEM…………………………………………………………7
2.1 Indoor climate aspects……………………………………………………....7
2.1.1 General indoor climate requirements………………………………7
2.1.2 Indoor climate problems associated with VAV systems………….8
2.2 Technical aspects……………………………………………….................12
2.3 Demands on a pressure independent VAV supply air diffuser………...17
2.4 Demands for other VAV system components……………………………18
2.5 Energy aspects……………………………………………………………...19
2.5.1 Electrical energy savings ...........................................................19
2.5.2 Thermal energy savings……………………………………………24
2.5.3 Energy savings for cooling…………………………………………26
3 FIELD MEASUREMENTS – TWO CASE STUDIES…………………...27

3.1 Case study 1 – EDIT-building at Chalmers University campus………..27
3.2 Case study 2 – Academicum and Anatomihöghuset at Sahlgrenska
Academy……………………………………………………………………..29
3.3 Conducted measurements in the case studies………………........…….32
3.3.1 Indoor climate measurements……………………………………..32
3.3.2 Energy consumption measurements..........................................33
3.4 Indoor climate evaluation – questionnaire…….…..…………………......33
4 LABORATORY MEASUREMENTS …………………………..…………37

4.1 The test room………………………………….........................................37
4.2 Conducted measurements in the laboratory.........................................39
5 RESULTS AND DISSCUSSION….. …………………………..…………43

5.1 Results from the case studies………………………...............................43
5.1.1 Thermal comfort measurements in measured rooms..………....43
5.1.2 Noise levels in measured rooms…………………………….........45
5.1.3 Indoor climate evaluation – questionnaire……..........………......46
vii
5.1.4 Energy consumption monitoring.……………...…..……..............51
5.1.5 Airflow rate and fan power….......…………………..…….............55
5.1.6 Specific Fan Power – SFP…………………..........................…...67
5.2 Results from the laboratory measurements……………………………...72
6 EVALUATION OF THE HEAT GAINS IN A VAV DUCT SYSTEM ….79

6.1 Mathematical calculation of the temperature rise in the duct system...79
6.2 Simulations and measurements with the duct system in the field.........86
7 CONCLUSIONS AND FUTURE RESEARCH ………………………….99

7.1 Conclusions………………………………………………………………….99
7.2 Future studies……………………………………………………….……..101
8 REFERENCES………………………………….…………………………103
APPENDICES……….…………………………………. ………………………...109
A Measurement instruments and measurement uncertainties………….111
A.1 Field measurements………………………………………………111
A.2 Laboratory measurements………………………………........….112
A.3 Measurement uncertainties………………………………………113
B Results from the laboratory measurements……………………….……117
C Questionnaire…………………………………………………………...…125
viii
NOMENCLATURE
Roman Letter Symbols
dA surface area of a duct element, [m2]

a1 inner perimeter area of a duct layer, [m]
a2 outer perimeter area of a duct layer, [m]
ai inner perimeter area of a duct, [m]
ao outer perimeter are of a duct, [m]
am logarithmic middle area of a duct layer, [m]
cp specific heat capacity of air, [J/(kg K)]
ci sensitivity coefficient
C& heat capacity of air, [W/°C]
DR Draught Rating, [%]
Dh hydraulic diameter of a duct, [m]
F F-ratio in F-test in statistical analysis
H0 zero hypothesis
K factor which depends on the airflow rate and on the device setting
k coverage factor
l thickness of a duct layer, [m]
li thickness of the insulation on a duct layer, [m]
lCD insulation thickness of a connection duct, [m]
lMD insulation thickness of a main duct, [m]
L duct length, [m]
Lc characteristic length, [m]
L0 perimeter area of a duct, [m]
m total number of adjustable parameters used in the model fitting
M& a air mass flow rate in a duct, [kg/s]
n fan speed, [revolutions per minute]
N number of input estimates xi on which the measurement output depends
neff total number of experimental points used in the model fitting
Nu Nusselts number
Nu trc Nusselt number for the transitional region with forced convection
Nulc Nusselt number for laminar forced convection
Nutc Nusselt number for turbulent forced convection
pmin minimum pressure value at given system point, given as a value above
the atmospheric pressure. [Pa]
pmax maximum pressure value at given system point, , given as a value
above the atmospheric pressure [Pa]
pSP pressure set point value for a pressure sensor, given as a value above
the atmospheric pressure [Pa]
P probability value in a statistical analysis
Pr Prandtl number
Re Reynolds number
ri inner radius of a duct, [m]
ro outer radius of a duct, [m]
yj jth repeated observation of randomly varying quantity y
qh annual specific heat energy needed for supply air, [MWh/year/(m3/s)]
ix
q& tr transmission heat loss, [W/m2]
dQ& c change in the heat capacity of the air flowing along the duct, [W]
Q& h power needed for central heating of air, [W]
Q& c power needed for central cooling, [W]
s (z ) standard deviation of a sample mean of n independent repeated values
for a quantity z
SD Standard deviation of air velocity, [m/s]
SFP Specific Fan Power for design conditions, [kW/(m3/s)]
SFPA Specific Fan Power for average airflow rates, [kW/(m3/s)]
SFPVAV Specific Fan Power for instant conditions in a VAV system,
[kW/(m3/s)]
tx steady state air temperature along a duct, [°C]
ta air temperature, [ºC]
tin inlet air temperature of a duct, [°C]
top operative temperature, [°C]
t0 outside temperature of a duct, [°C]
tpr plane radiant temperature, [°C]
trec air temperature after the heat recovery, [°C]
troom room air temperature, [°C]
tr mean radiant temperature, [°C]
tsupply supply air temperature, [°C]
Tcal calculated T-value for the student’s T-test
Tcal critical T-value for the student’s T-test
Tu turbulence intensity, [%]
ul linear thermal transmittance of a duct, [W/(m .K)]
U thermal transmittance between the air in the duct and outside
temperature, [W/(m2 .K)]
u ( xi ) standard uncertainty associated with the estimated value of each input
quantity xi.
uc(y) combined standard uncertainty of output estimate y.
ue expanded uncertainty of output estimate y
v air velocity of the airflow in the duct, m/s
va local mean air velocity, [m/s]
var relative air velocity, [m/s]
vd supply air velocity from the diffuser, [m/s]
V& airflow rate, [m3/s]
V&VAV instant airflow rate in a VAV-system, [m3/s]
V&min minimum airflow rate, [m3/s]
V& design design airflow rate, [m3/s]
W& t total power input to a fan motor, [W]
W& t design total power input to a fan motor at the design airflow rate, [W]
wi weights of each experimental point
xi measured values of the independent variable x
yi measured values of the dependent variable y
yijklm a single observation of a variable (air velocity) in a mathematical model
y general mean of a variable (air velocity) in a mathematical model
x
z arithmetic mean of n repeated observations of randomly varying
quantity q
zj jth repeated observation of randomly varying quantity z
Greek symbols
αi convective heat transfer coefficient between the air and inner duct
surface, [W/(m2°C)]
αo heat transfer coefficient between the outside air and outside duct
surface, [W/(m2 .K)]
αj room level effect in a mathematical model
βi room point effect in a mathematical model
τk ceiling effect in a mathematical model
γl airflow effect in a mathematical model
λd thermal conductivity of a duct layer, [W/(m .K)]
λi thermal conductivity of a insulation layer, [W/(m .K)]
λi thermal conductivity of metal, [W/(m .K)]
λ air thermal conductivity of air, [W/(m.K)]
∆h enthalpy decrease in the system, [kJ/kg]
∆pAHU pressure drops in the air handling unit.
∆ptot fan total pressure, [Pa]
∆pfmin fan pressure at minimum airflow rate, [Pa]
∆pdesign fan pressure at design airflow rate, [Pa]
∆pS fan pressure at static pressure set point level, [Pa]
∆pmin system pressure drop at minimum airflow rate, [Pa]
∆pm measured pressure difference in the measuring device, Pa
∆pduct pressure drops in the duct system, [Pa]
∆t1 temperature change in the duct, [°C]
∆t2 temperature difference between the inlet air temperature and outside
temperature, [°C]
υ kinematic viscosity of air, m2/s
ηtot total efficiency of the fan system
ρ density of air, [kg/m3]
ηT2 temperature efficiency of the heat recovery system
χ the sum of squares of the deviations of the theoretical curve from the
experimental points
εijkl error in a mathematical model
Abbreviations
ACH air changes per hour

AHU Air handling unit
BTA brutto area of the building
CAV Constant air volume flow
CC1 Frequency inverter with PID control
CC2 Control center
CD connection duct
F Duct filter
xi
FM Flow measurement
HC Heating coil
HVAC Heating, ventilation, air conditioning
IAQ Indoor air quality
LOA premises area of the building
PPD predicted percentage of dissatisfied people
PS Pressure sensor
SC Suspended ceiling
SD Supply air diffuser
SA Sound attenuator
SF Supply air fan
SP static pressure sensor
T Thermometer
TS Temperature sensor
VAV Variable air volume flow
xii
1 INTRODUCTION
1.1 Background
1.1.1 General
In commercial buildings the need of energy for the building services systems often
accounts for a substantial part of the total energy consumption. In Swedish office
buildings the electrical energy needed for these systems may account up to 50% of the
total electricity use [48]. The need of heat for supply air heating also sometimes
constitutes a considerable part of the total need of heat. Although the annual heat
energy use for office buildings has decreased considerably due to better design of the
building shell and air-to-air heat recovery, the amount of total electrical energy needed
for commercial buildings has not decreased. Figure 1.1 gives an example of measured
energy consumption in selected office buildings Sweden from different construction
years.
Higher heat loads in rooms, caused by an increased use of electrically powered office
equipment, increased glazing, denser use of the floor area, etc., have resulted in higher
demands for ventilation and cooling, which have lead to an increase of electricity use
in general. Ways to use this energy more efficiently have become a goal of many
researches in the field of heating, ventilating and air conditioning (HVAC) systems.
Energy conservation measures should be looked for both when retrofitting existing
buildings and buildings new ones.
[kWh/m /year]
300 Electricity Heating

,
2 2
use [kWh/m
250
200
energyuse,
150
Specificenergy
100
50
Specific
0
1960 1970 1973 1976 1979 1981 1985 1988 1990
Year of construction
Figure 1.1. Relationship between year of construction and the use of delivered
energy use for a selection of Swedish office buildings [48].
The majority of Swedish existing office buildings, constructed before about 1985,
have centralized constant air volume flow (CAV) air conditioning systems with the
primary function to keep the room temperature at an acceptable level summertime.
Nowadays air based air conditioning systems in office buildings have variable air
volume flow (VAV).
1
1.1.2 CAV and VAV
In the older CAV equipped office buildings the airflow rate is usually decided by the
heat load that has to be mastered. However, the heat loads in the rooms are only
occasionally at the peak level. They vary with time of the day and time of the year and
also with the location of the room/zone in the building. In addition, the activities in the
rooms result in a change of heat loads. When the air volume flow is kept constant
(CAV system), the airflow rates will be high most of the time. Even if the systems
usually are provided with air-to-air heat recovery and are in operation mainly during
working hours, the need of energy for air heating and air distribution is quite high.
Moreover, when the cooling capacity of the supply air is almost constant, as it is in a
CAV system, there is a need for additional heat supply to the room during periods
when the internal heat generation is low (see figure 1.2).
If the air flow is continuously adopted to the actual heat load, considerable energy
savings can be achieved [27, 37, 62]. The average air volume flow will be lower and the
energy needed for air distribution and supply air heating is reduced [19, 28, 39].
Additional heating in the room can be avoided, since the cooling capacity of supply air
is controlled according to the demand of the room or zone (see fig. 1.2).
2
2
Heat generation- Heat losses, W/m
CAV cooling capacity

Heat generation- Heat losses, W/m
VAV cooling capacity
Ventilation
heat losses
Heat loads Heat loads
Transmission Transmission
heat losses heat losses
0 3 6 9 12 15 18 21 24 0 3 6 9 12 15 18 21 24
Time of Day Time of Day
Figure 1.2. Typical variations of the heat generation and heat losses in an office
building. In a CAV system the heat losses caused by cold supply air
must be covered by room heating. In a VAV system this extra heat is
not needed
1.1.3 The supply air temperature

Additional energy savings can be achieved if the supply air temperature to the room
can be decreased. One cause of the high need of heat energy in a CAV system is that
the outlet air devices quite often need rather high supply air temperature in order to
avoid disturbing draught in the room. This is the case e.g. when displacement
ventilation is installed in small rooms, which was quite frequent in Sweden when
existing offices were rebuilt in the early 1990’s. For an acceptable thermal comfort in
these rooms, often a supply air temperature of at least +19°C is needed in practice.
Decreasing the supply air temperature will not only give energy savings in supply air
heating, but it will also improve the cooling capacity of the air supplied and lead to a
2
better control of the room temperature. The average airflow rates needed for cooling
are reduced even more.
Figure 1.3 shows the airflow rate needed to cover the cooling needs in the room [49].
The cooling capacity of the supply air depends on the temperature difference between
the room air and the supply air. As it can be seen from the figure, with a higher
temperature difference between the room and the supply air, the lower the required
airflow rate needed for the cooling of room will be.
Supply air
cooling Temperature
capacity difference troom-tsupply
W
10°C 8°C 6°C
300
200 4°C
100 2°C
0
0 10 20 30 40
Airflow rate (l/s)
Figure 1.3 The airflow rate needed for achieving the required cooling capacity
depending on the temperature difference between the room air and the
supply air [49]. Example: with the airflow rate of 30 l/s and with the
temperature difference of 6°C, a cooling capacity of 220W is achieved.
However, with the same cooling need, this airflow rate can be decreased
almost by half when the temperature difference troom-tsupply is increased
to 10°C.
In temperate climates extra advantages can be taken from free-cooling when the
outdoor temperatures are lower than the supply air temperature [50].
1.2 Objectives and scope

The aim of this work is to:
• clarify the prerequisites for well functioning VAV systems in office buildings;
• study the possibilities to meet these prerequisites by laboratory tests;
• examine the VAV system function in buildings in full scale operation.
A special focus is that the system solutions should be uncomplicated and they should
be possible to implement in existing buildings as well as in new ones.
The first demand that has to be set on the system is that the requirements on the indoor
climate parameters are fulfilled. Consequently, the first step of the study is to define
indoor climate demands. Naturally the system should not create any problems with
thermal comfort or disturbing noise in the occupied rooms.
3
Based on the requirements on the indoor climate, the demands on the components of
the system and the system itself are defined.
One example of a technical solution, which seemed to have properties that fulfil the
demands is tested in laboratory as well as tested and analysed in plants in function,
with focus on indoor climate and the need of energy.
1.3 Methodology
A special focus in this work is to study VAV system solutions that are not so
complicated from a system control point of view. One criteria of an uncomplicated
system is that the number of controlling components is minimized. That means, e.g.,
that operating dampers in the air distribution system should be avoided as far as
possible. This is especially important when existing CAV systems are changed to
VAV, as it may be complicated to install new dampers in an existing duct system.
Adapting VAV in office buildings without having to install active control dampers in
the duct system is one way of building up a simple VAV system. The active control
damper is defined as an actuator controlled damper. However, to build up a system
without active dampers some special demands must be set upon the VAV variable
supply air diffusers. These demands have been analysed in this study.
The methodology comprises three different ways of studies:

• case studies
• laboratory experiments
• mathematical modelling
The aim of the case studies was to measure the function of the VAV system in the
field, in order to see how the described VAV system without active control dampers
can be applied and what the demands and requirements for the system design and
components are. The tests are made both in a new office building and in an existing
office building, where an old CAV system was rebuilt to VAV.
The selected case study objects give a possibility to analyse the ways of adapting the
VAV system in both new and old buildings. In a new building the system can be
adopted for VAV from the beginning. In existing buildings extensive measures in the
system may be difficult. Consequently, requirements for the system components and
their application requirements for both new buildings and existing buildings have been
analysed.
Since well-controlled parameter studies are difficult to conduct in the field, a full size
office room was built up in laboratory in order to test the function of a variable supply
air diffuser. The aim of the laboratory studies was to see how the requirements from
the indoor climate point of view are met under different flow conditions.
The last part of the methodology consists of mathematical calculations. Here the
implication of varying airflow rates on the temperature rise in the ducts has been
evaluated and also verified in the field. The cold air streams in ducts are influenced by
heat gains due to the temperature difference between the supply air in the duct and the
4
room temperature. That influences the cooling capacity of the air supplied to the room.
A simple mathematical modelling has been conducted to see how much the airflow
rates can be decreased, without having strong implication on the temperature control
in the room. Also the means of decreasing the heat gains in the cooled air system have
been evaluated.
1.4 Outline of the thesis

The introductory part, chapter 1, gives the background and the aims of the current
study. The description of the chosen methodology is also included here.
In chapter 2, the prerequisites for a VAV system without active control dampers in the
ducts are discussed. The indoor climate aspects are analysed and possible technical
solutions and requirements for system components for fulfilling the set demands are
presented. The discussion of the theoretical energy consumption aspects of a VAV
system is included here.
Chapter 3 then describes the field measurements carried out, while chapter 4 accounts
for the studies in a well-defined laboratory set up.
The results from the case studies and laboratory measurements are discussed and some
conclusions made in chapter 5.
The heat gains in a VAV duct system are analysed and mathematically modelled in
chapter 6.
The final conclusions and recommendations for future research are given in chapter 7.
Appendix A describes the measurement instruments used in the field and laboratory
measurements. Also the description of the measurement uncertainties is presented
here.
Appendix B presents some of the results from the laboratory measurements as a

complement to the discussion given in chapter 5.
Appendix C shows the indoor climate questionnaire that was used in the case studies is
given
1.5 Literature review

Variable air volume systems became internationally installed in1980s, when
acceptably well functioning technical solutions became available. Even though in
Sweden the VAV systems have not been as common as in the US and in the UK, this
system type has been installed here in some extent as well. The increasing energy
prices also supported energy saving solutions. Ever since a number of studies has been
conducted regarding VAV systems and their applications. A review of published
references can be found in the bibliography of the book “VAV Air Conditioning
Systems” [55].
5
The research done so far concentrates mainly on achieved energy savings, problems
occurring with VAV and analysis of different types of VAV configurations and their
control techniques. The main advantages of VAV compared to CAV have been
analysed and studied a lot from both energy savings aspects and savings in initial costs
[14, 28, 39, 45, 46, 50]
.
However, some common problems may occur when operating VAV systems. Such
problems have been analysed in a number of reports during the last decade. The main
problems concern the indoor climate in rooms [53, 55, 61]. Operational and control
problems are also common [14, 15]. After analysing the sources of these faults it has
been indicated that the reasons could be improper design of the system components by
manufactures, as well as wrong design and installations by building contactors [29, 42].
Nevertheless, because of the fast development of the system components of VAV
systems during the last 10 years the problems occurring have been reported to be due
to improper design, application or operation of the systems [42, 43, 66].
The majority of conducted studies concentrate on VAV systems with active control
dampers in the duct system and the use of pressure dependent supply air devices. This
configuration has been so far the most common according to different guide books [27,
55]
.
There appears to be a lack of studies that analyse the suitable applications of VAV
systems when retrofitting existing commercial buildings. In the studies concerning
renovation of existing buildings [37, 62], the rebuilding from CAV to VAV focused
mainly on variable frequency drives and pressure control in the duct system. However,
it has been shown that problems with indoor climate may occur due to poor adaptation
of the rest of the system to variable air volume flow [53, 55].
6
2 VAV AS A SYSTEM
2.1 Indoor climate aspects
2.1.1 General indoor climate requirements
The main goal of heating, ventilating and air-conditioning (HVAC) systems is to
create a good indoor climate for the building occupants, with reference to indoor air
quality (IAQ), thermal comfort and acoustic environment. These parameters may have
a significant influence on productivity and work efficiency [23, 65].
The requirements of the occupants on air quality is firstly that there should not be a
health risk and secondly that the air should be perceived as fresh and pleasant [11]. This
is achieved by supplying enough outdoor air to the occupied zone in order to keep the
pollutants from materials and building occupants at an acceptable level. Design
guidelines commonly set the minimum levels of outdoor airflow rates that should be
assured when ventilating the spaces.
The requirements for thermal comfort define the conditions, where the building
occupants would perceive the thermal environment to be most satisfying. It states the
conditions where the human body would be in thermal balance with the surroundings.
The human response to the thermal environment depends on parameters such as air
temperature, mean radiant temperature, air velocity and air humidity. It also depends
on personal factors such as the occupants’ physical activity and the thermal resistance
of the clothing.
A person does not feel the room temperature directly. What is felt is the heat loss from
the body. Therefore the parameters measured, when evaluating thermal indoor climate,
should be the ones that affect this energy loss. In general the amount of heat given off
by a human body through radiation is approximately the same as the amount of heat
given off by convection. Therefore measuring only air temperature is not a good
indication of the thermal environment and a radiation effect from the surrounding
surfaces should be also considered. This can be taken into account by measuring mean
radiant temperature, which is a measure of the combined effects of temperatures of
surrounding surfaces in the room. It is defined as the uniform temperature of an
imaginary black enclosure which would result in the same heat loss by radiation from
the person as the actual non-uniform enclosure [3].
Nevertheless, measuring the mean radiant temperature can be very time consuming,
since the temperatures of all surrounding surfaces should be measured. Many of the
thermal comfort standards [5, 30] and design guidelines [11] recommend measuring a
temperature that is representative for the operative temperature. The operative
temperature combines the air temperature and the mean radiant temperature into one
numerical quantity. It can be characterized as the uniform temperature of an imaginary
black enclosure in which an occupant would exchange the same amount of heat by
radiation and convention as in an actual non-uniform environment [5].
Existing comfort standards, such as ISO 7730 [30] and ASHRAE Standard 55-2004[5],
give the combinations of thermal and personal factors with which at least 80% of the
7
building occupants are expected to be satisfied. According to these standards a human
response to the thermal environment can expressed to by the predicted mean vote
(PMV) and predicted percentage of dissatisfied (PPD) indices, which predict the
percentage of the occupants feeling too warm or too cool for the body as whole. The
human response is also expressed by the percentage of occupants predicted to feel
dissatisfied due to different types of local thermal discomfort. The local thermal
discomfort may be caused by draught, by abnormally high vertical temperature
difference, too cold or too warm floor temperature or by too high radiant temperature
asymmetry. These thermal comfort models are based on a large number of laboratory
studies in which subjects were asked to evaluate their comfort in steady-state
environments.
The design guideline of indoor environment CR 1752 [11], which is based ISO 7730
standard [30], divides the desired thermal environments into three different categories
A, B and C. The categories are depending on the number of people who would be
dissatisfied due to their thermal environment. The highest category A refers to less
than 6% dissatisfied for the body as a whole (PPD) and less than 15% dissatisfied for
local discomfort. For achieving these conditions for thermal environment in a cellular
office the operative temperature should be kept in the range of 24,5°C ± 1,0°C during
summertime and 22,0°C ± 1,0 °C during wintertime. According to the same design
guideline [11] the limitation for the mean air velocity in the occupied zone is 0,18 m/s
for the summer period and 0,15 m/s for the winter period. The same limits for the air
velocity are also given in Swedish design guidelines BFS 1993:57 [8] and R1 [52].
The requirements for the acoustic environment define the maximum acceptable noise
levels generated by different installations, including HVAC systems. These noise
levels should be kept in a level that will not cause a significant nuisance or health
effects for building occupants and will not disturb work. In general the air should be
distributed and removed as silent as possible. The major sources for noise generation
in HVAC systems are the fan, the duct system with dampers and the diffusers. It is
essential to evaluate these noise sources already in the design process and preventive
measures should be taken into account in order to reduce them. The Swedish Standard
for building acoustics [57] classifies the maximum acceptable noise levels generated by
different installations into four different groups A, B, C and D. According to the
highest noise requirements for the room (class A/B), the maximum acceptable sound
pressure level from installations in an office room is 35 dB(A).
2.1.2 Indoor climate problems associated with VAV

systems
The indoor climate aspects described in previous chapter form a basis for design
criteria for all of the HVAC system and not only for variable air volume (VAV)
systems. However, there are some common indoor climate problems that are
associated especially with VAV air supply and pointed out in previous research. These
common complaints are the following.
Insufficient supply of outdoor air
This is mainly a problem in the countries where the main practice is to use circulated
exhaust air in an “outside air economizer cycle” for a heat/cooling recovery system.
8
This is used widely in US, Canada and also in UK in Europe. A lot of research has
been done on the VAV systems that use return air, in order to determine the amount of
outdoor air in the supply air needed both for proper indoor air quality (IAQ) and to
achieve energy savings [2, 25, 43, 50, 67]. However, in the Northern countries air-to-air heat
exchangers for the heat recovery are used and the VAV systems usually operate with
100% outdoor air [19, 20]. Therefore a lack of outside air will not be a problem here,
presupposed that the minimum design air volume flow is properly chosen.
“Stagnant air”
According to previous studies, this complaint may occur in circumstances where

minimum airflow rates from the supply air device do not guarantee enough air motion
[10, 29, 42, 61]
. This condition has been stated to predominate during heating seasons, if
the temperature of the supply air is increased to the room temperature level or higher.
This is more common in US, where the VAV system may be used both for heating and
cooling the building.
The problem has also been addressed to improper VAV box control. The required
supply airflow rates may not be assured and the space loads are not met, causing the
room temperature to rise. Therefore, the feeling of too little air motion in the room
may be caused by a high room temperature.
There are no guidelines that report the minimum air motion required for comfort.
However, it is stated that increasing the air motion may increase comfort if the air
temperature is too high [5]. Moreover, according to comfort equations in the comfort
standards [5, 30], the optimum comfort will be achieved only with low air speed and
proper temperatures. Therefore adjusting room temperatures may be even more
effective than adjusting air speeds to achieve thermal comfort. This indication has also
been reported from different studies [60, 61].
Nevertheless, the air circulation in a room is a very complex process and lots of
studies have been carried out in order to understand the behaviour of room ventilation
[41]
. The air movement in a room with a certain supply air device can be considered as
function of two factors, air volume and temperature difference between the supply air
and the room air. At high airflows, the predominant factor is the air quantity and the
resultant discharge velocity from the diffuser. However, with lower airflow rates the
temperature difference becomes important and will be a strong force in determining
the air movement in the room. Decreasing this temperature difference results in higher
air motion in the room [29]. Therefore, the problem of “stagnant air” is unlikely to
occur if the supply air temperature is kept lower than the room temperature.
Draught
Draught has been reported to be one of the common problems associated with VAV
air distribution [10, 42, 55]. It is defined as an unwanted local cooling of the human body
caused by air movement in the room. The air movement affects the thermal balance of
the human body by increasing the convective heat loss of the body and increasing the
evaporation of skin moisture. As a result, a person feels thermally uncomfortable.
The amount of heat loss from the skin caused by draught is not only dependent on the
air temperature and the local air velocities in the occupied zone. It also depends on the
9
fluctuation of the air velocity, described as turbulence intensity [24]. Draught
complaints may occur even in circumstances when measured air velocities are lower
than prescribed in existing standards.
A high turbulent airflow is felt to be more annoying than a low turbulent airflow, even
though they result in the same heat loss. This is due to a fluctuation in skin
temperature that causes discomfort. The air velocity fluctuations with a frequency of
0,5 Hz has been reported to be the most uncomfortable, while frequencies above 2 Hz
are not felt [9].
The draught is usually felt on the unclothed parts of the body like face, hands and
lower legs (local discomfort). Therefore the air velocity and air temperature
measurements for establishing local thermal discomfort due to draught should be
carried out at the heights of head, hands and ankles [9].
The problem with draught in rooms with VAV systems is normally caused by the use
of improper diffuser or too low supply air temperatures. With decreasing airflow the
air velocities in the discharge opening of the supply air outlets will sink if the
discharge area is constant. Due to that cold air may drop to the occupied space,
causing draught. Therefore it is important that the air velocities of the supply air path
in different flow conditions are kept on the same level, which should be considered
when choosing a supply air device for variable airflow.
Estimation of draught rating (DR) is one method to evaluate the draught risk in
existing rooms. It expresses the percentage of people predicted to be bothered by
draught [30] and is based on studies on people at light sedentary activity, with an
overall thermal sensation for the whole body close to neutral. The sensation of draught
is lower at activities higher than sedentary and for people feeling warmer than neutral.
For calculating the draught rating (DR) the air temperature and local air velocities
should be measured in a room point where the draught rate is evaluated. Most
commonly the measurements are carried out next to the working place.
The draught rate can be calculated by the equation:
DR = (34 − t a )(va − 0,05) 0,62 (0,37 ⋅ va ⋅ Tu + 3,14) [%] (2.1)
Where,
DR Draught Rating, [%]
ta air temperature, [ºC]
va local mean air velocity, [m/s]
Tu turbulence intensity, [%]
The turbulence intensity Tu considers the fluctuation of the air velocity and can be
calculated according to equation 2.2:
SD
Tu = ⋅100 [%] (2.2)
va
10
Where,
SD Standard deviation of air velocity, [m/s]
For calculating SD and va a period of 3 minutes is used. The transducers used for
draught rating measurements must be able to measure the air velocities down to 0,05
m/s and the fluctuations up to 2 Hz and must be unaffected by the direction of the
airflow [9].
The draught rating (DR) itself cannot be evaluated in the design process, but can be
determined in buildings in operation. However, the design criteria define the room
temperatures and air velocities that should be assured for achieving good thermal
comfort in the occupied space and in order to avoid draught. In ISO 7730 standard [30]
the elimination of draft is addressed by placing limits on the allowable mean air
velocity as a function of air temperature and turbulence intensity. As an example, the
draft risk data represented by 15% dissatisfaction curves would restrict the mean air
speed to 0.12 m/s at 20°C and 0.2 m/s at 26°C. This applies for an assumed turbulence
intensity of 40%, which is typical for indoor office environments.
Disturbing sound
Variations in airflow rates lead to variations in the static pressure in the ducts. The
decrease in airflow rates at VAV terminal units/diffusers causes changes in system
characteristics. As a consequence an increase in fan total pressure will occur if the fan
speed is kept constant. Moreover, as the airflow rate is reduced, the duct system
pressure drop is reduced. The combined effect of these two factors will result in higher
static pressure at supply air terminal units/devices as airflow is cut down. The excess
of pressure must be simply throttled off somewhere in the system. This can be done at
the terminal units or at pressure reducing dampers. As a consequence, the noise levels
of the devices will increase by the amount stated in the rating tables of the particular
devices [14].
If the increase in static pressure and noise levels in the system is within acceptable
limits, there will be no need for control of pressures in the central air handling system.
Otherwise the static pressure control should be applied to the system. This is
commonly done by controlling the fan speed.
Nevertheless, despite the pressure control, there will still be some pressure variations
throughout the system between the air terminal units and the location of the static
pressure sensors. Therefore, careful consideration should be given when selecting the
flow controlling units/devices. Here the throttling action of the damper is most
commonly the cause of extra noise. Moreover, an improper selection of VAV boxes
size can result in a high static pressure drop across the box when it is selected too
small [42].
2.2 Technical aspects

A VAV system most commonly consists of the following main components:
• an air conditioning system together with filters, an air-to-air heat exchanger
and coils for supply air heating and cooling (if needed)
11
• a fan system with the equipment for the pressure control (frequency converters
and pressure sensors for keeping constant pressure in the duct system)
• an air distribution system with flow control units.
Here the last components may be the most demanding part of the VAV system both
from technical and economical point of view. It is a simple task to add flow control
properties to a central fan system by just installing frequency inverters to a fan system
and adding pressure control to the system. However, it is much more complicated to
adapt the rest of the system to variable airflow. The selected components on the air
distribution system side can have a big influence to VAV system characteristics and
therefore will be further analysed.
In a basic CAV system the constant volume flow rate is controlled and set to the
desired flow with balancing dampers on each duct branch. Here the balancing dampers
are defined as manual operated flow adjustment dampers. The pressure drops across
the supply air devices are selected so that when combined with the pressure drops
across the balancing dampers they give the right flow distribution. Balancing dampers
are also installed on the main ducts for each zone in order to set the airflow rate
between zones/floors (see figure 2.1). Quite a number of balancing dampers are
needed in order to assure the required airflow rates to rooms/zones.
CAV
Balancing damper
Figure 2.1. A configuration of a constant air volume system (CAV). The balancing
dampers for flow adjustment are needed both on main ducts to each
zone as well to each branch duct.
In a VAV system, these manually set balancing dampers are not always needed, since
the airflow rate will be controlled and set to the required level by variable VAV supply
air diffuser or airflow terminal units (VAV boxes). However, the number of other
controlling equipment needed for the VAV system operation depends to a great extent
on how the pressure unbalance in a system, caused by the initial design and partial
load operation, will be absorbed.
The variation in airflow rates leads to a variation in static pressure in the system. The
decrease in airflow rates at VAV terminal units causes changes in system
characteristics, as shown in figure 2.2. As a consequence an increase in fan total
pressure will occur if the fan speed is kept constant. Due to the increase in fan total
pressure and reduced pressure drops in the duct system as the airflow is reduced, an
excessive increase in static pressure levels may occur at the supply air devices, unless
pressure control is applied.
12
Total
pressure
rise, ∆ptot VAV devices closing
∆pfmin
System curve
∆pdesign
Fan curve
V&min V&design Airflow rate,V&

Figure 2.2. Change in fan total pressure with decreased airflow rates and constant
fan speed in a fan with backward curved blades. The closing action of
VAV terminal units for decreasing airflow rates causes changes in the
system characteristics. As a consequence an increase in fan total
pressure will occur. In the figure: ∆pdesign - fan pressure rise at design
airflow rate; ∆pfmin – fan pressure rise at minimum airflow rate.
The requirement for duct pressure control in the central system is generally dictated by
the amount of air volume reduction, overall acoustical requirements and economy of
the system costs [14, 28, 45]. If the increase in pressure is managed by the airflow
controlling devices and noise levels in the system are within acceptable limits, there
will be no need for additional control of pressures in the central air handling system.
The excess in pressures will simply be throttled off somewhere in the system, which
can take place in flow terminal units or at pressure-reducing dampers. However, this
inevitably results in wasteful fan energy consumption, since the system actually needs
less fan total pressure with reducing airflow rates.
A common practice for static pressure control in the system is to control the fan speed.
The fan speed will be reduced with reduced airflow rates in the system, as shown in
figure 2.3. The static pressure is kept on a level that can assure a proper work of the
flow control devices at maximum airflow rates.
However, there is still no consensus on where the pressure sensor should be installed
in the duct system from an energy point of view. The question is if the right place is to
install it in the beginning, in the middle or in the end of the duct system. Here the
VAV system with the pressure sensor installed in the beginning of the duct system will
be discussed.
13
Total
pressure
rise, ∆ptot
Fan speed n1
n2 System curve
∆pdesign n3
n4
n5
∆pS
V&min V&design Airflow rate,V&

Figure 2.3. A static pressure control in a VAV system by controlling the fan speed.
The pressure ∆pS is kept on a level that can assure a proper work of the
flow control devices at the design airflow rate.
Nevertheless, control of the static pressure at the fan outlet compensates only the
system pressure changes produced by the fan characteristics. The pressure rise due to
the resistance changes in the duct system, that must be throttled off somewhere in the
air distribution system with decreased airflow rates, can still be relatively high. This
can be the case for example in a VAV system where the airflow rates vary in a
considerably big range. With minimum airflow conditions the pressure drops along the
duct system can be very low and the static pressure levels at the supply air devices can
be high compared to maximum flow conditions (see figure 2.4).
VAV SP
20-120 Pa
Pressure in
2 3
the system,
Pa
150Pa
pSP
pminflow = 120Pa
pmaxflow = 20-30Pa
1 2 3 Distance
Figure 2.4. Example of pressure changes in a VAV system with maximum and
minimum airflow rates. In the figure: pminflow - pressure value at system
point 3 with minimum airflow; pmaxflow - pressure value at system point
3 with maximum airflow, pSP - pressure set point value for the pressure
sensor SP.
14
Depending on how the change in pressures will be managed in the system, the airflow
control diffusers/units are classified as pressure independent or pressure dependent
diffusers/units. A pressure independent VAV diffuser/unit can be defined as a device
where the airflow control is not disturbed by the fluctuations of the static pressure at
the inlet of the device [55]. A preliminary requirement is that these fluctuations of
pressure remain within the maximum and minimum limits specified by the
manufacturer of the device. Keeping the required static pressure in the main ducts after
the fan system can control these limits.
A pressure dependent VAV diffuser/unit is a device where the airflow control is

depending on the static pressure at the inlet side of the device. The airflow will vary
with the fluctuations in static pressure before the device and a separate provision must
be made to limit these fluctuations to an acceptable level. An example of a VAV
system with pressure dependent flow control devices is given in figure 2.5. Here the
airflow to the rooms/zones is controlled with the supply air diffusers with variable
discharge area. The flow is determined based on the device's opening (between 0 to
100%) and the underlying constant static pressure. Therefore, in order to assure proper
work of the variable VAV diffusers, the level of a pressure in the inlet side should be
strictly controlled. This is assured with installing active control dampers on the ducts
serving different zones for maintaining constant pressure, for example at approx. 50
Pa. Here the active control damper is defined as an actuator controlled damper.
VAV SP
SP
Active control
damper
SP 20-50Pa
Pressure dependent
VAV supply air diffuser
Figure 2.5. A VAV system with pressure dependent airflow control devices
(variable supply air diffusers). The pressure sensors SP1 and SP2 set
point values are kept on a level that can assure a proper work of a
pressure dependent VAV supply air diffusers at the design airflow rate.
However, all extra controlling equipment installed to the system makes the system in
general more complicated and costly. This can be especially a case when adopting
VAV in existing buildings, where extensive refurbishment is not always possible and
the active control dampers are not easy to install in the duct system.
These active control dampers can be avoided if the airflow control devices are
pressure independent. An example of this kind of system is presented on figure 2.6.
Here the airflow is controlled with the air terminal unit, most commonly called as a
VAV box. It consists of a control section with the airflow rate control damper and a
15
sound attenuator after the control section for decreasing the noise generated by the
damper.
VAV SP
20-120Pa
Pressure independent air

terminal unit (VAV box)
Figure 2.6. A VAV system with pressure independent supply air terminal units
(VAV box). The pressure sensor SP set point value is kept on a level
that can assure a proper work of a pressure independent VAV terminal
unit at the design airflow rate.
Even if the air terminal unit is supplied with the sound attenuator noise problems may
occur with high pressure drops over the device. This is referred to be one of the most
common problems associated with a VAV boxes [42]. Also the lower limits of airflow
rates are kept relatively high due to the difficulties to control very low airflow rates
and if a risk of draught should be avoided. In common installations the outdoor air to
the room is not supplied directly from the pressure independent VAV supply air
terminal unit and separate supply air diffusers connected to the VAV box are needed.
The draught risk may especially occur when using VAV boxes together with supply
air diffusers with constant discharge area, such as CAV supply air devices.
When technical solutions are selected, simplicity, intelligibility and tolerance against
deviations in operating conditions should always be looked for. In this current project
the possibility of building up the VAV system as simple as possible without
jeopardizing the demands for indoor climate is studied. Reducing the number of active
control dampers by building up a pressure independent system contributes to that.
Moreover, if the airflow control and room air distribution components can be added
together into one unit, the number of required system components can be reduced even
more. An example would be building up a system with pressure independent variable
supply air diffusers. However, in order to assure good indoor climate with this kind of
system solution, there are certain requirements that a pressure independent variable
supply air diffuser must fulfil. These demands will be described and analysed in the
following chapter.
16
2.3 Demands on a pressure independent VAV
supply air diffuser
A variable supply air diffuser is here defined as a supply air diffuser, which changes
its outlet configuration automatically when controlling the supplied airflow rate. The
parameters that control the amount of air supplied to the room can be temperature,
CO2 or presence in the room
The work principle of a common variable supply air diffuser is the following. The air
is supplied between a controlling plate with a distance that varies according to the
airflow rate needed (see figure 2.7). The position of the plate is controlled by a
traversing motor, which gets impulses from the controlling sensor locating in the
room. The controlling sensor can be a temperature sensor, a CO2 sensor, a presence
sensor or a combination of some of these. The flow is determined based on the
diffuser's opening, which is between 0 to 100% and the underlying constant static
pressure.
Supply duct
connection
Traversing motor
Controlling
plate
Figure 2.7. A scheme of a variable supply air diffuser. The air is supplied between
a controlling plate with a distance that varies according to the airflow
needed.
The basic idea behind the variable supply air diffuser is to maintain a constant velocity
of the supply air stream discharged from the diffuser with a decreasing supply airflow
rate. The airflow pattern issuing from the diffuser, which discharges the air
horizontally across the ceiling, has a natural tendency to cling to the surface. This
effect, called Coanda effect, is apparent as a negative pressure or suction that pulls
each layer of the air in the jet towards the ceiling. The cooled supply air is naturally
denser than the room air and affected by a negative buoyancy force that tends to cause
the supply air stream to fall away from the ceiling. If the negative buoyancy outweighs
the strength of the Coanda effect, a dumping of the supply air jet will occur and may
result a sensation of draught in the occupied zone [55].
The strength of the suction force tending to keep the air stream attached to the ceiling
is generally regarded as being proportional to the square of the discharge velocity. If
the discharge area of the diffuser remains constant, the velocity of the supply air
stream falls in direct proportion to the reduced airflow rate, resulting a risk of the
supply air jet falling away from the ceiling.
17
This risk has been considered in a variable supply air diffuser’s work, which controls
the diffuser openings in a way that relatively constant air velocity range is maintained
and dumping will be avoided.
The commonly used variable supply air diffusers are pressure dependent. In order to
assure a proper work of this type of a supply air device, a pressure in a range of 20-
50Pa should be maintained on the inlet side of the diffuser. This requirement is
achieved by keeping constant pressure in the branch duct by active control dampers.
In order to build up the VAV system without these active control dampers and to use
variable supply air diffusers for airflow control in rooms, the variable supply air
device must be independent of the pressure changes in the system. This means that it
should be able to manage the pressure variations in the system without any noise
problems and any interference to the supply airflow rate.
In general the pressure independent variable supply air diffuser must have following
variable air volume flow properties:
• Demand based airflow control by room temperature, by room air quality, by

presence detection or by a combination of these should be possible.
• The supply airflow rate should be independent of the pressure variations in the
duct.
• A high pressure difference over the device, at least 100 Pa, must be possible to
manage without noise problems in an airflow range from minimum to
maximum.
• The diffuser should manage to control the airflow in a wide airflow range,
from 5-100% of the design airflow rate
• The supply airflow pattern to the room must be stable and independent of the
supply airflow rate in order to assure good air movement and avoid the cold
supply air dropping into occupied zone.
• It should be possible to supply air with a low temperature, +15°C or lower,

without risk of disturbing draught
• The cost must be acceptable from a life cycle point of view.
2.4 Demands for other VAV system components

For a stable operation of a VAV system there are some general requirements for
system components and control system to fulfil.
It is important to maintain the constant pressure in the system. When the outlets will
absorb the greater part of pressure unbalance caused by partial load conditions, then
the main function of static pressure control in the central unit is to prevent excessive
pressure build up in the duct system caused by the fan characteristics.
18
A balance between the supply and return air should be assured with variations in
supply air for a VAV system proper work. Balancing of the system can be done by
measuring supply and exhaust airflow rates in the main duct lines of the zones and
adjusting the exhaust airflow rates respectively with the damper installed on the main
exhaust air ducts.
There are also some points to consider when designing the duct system. With the
decreased airflow rates in cooled supply air ducts the heat gains can have a significant
effect to the supply air temperature. The heat transferred through the duct system is
dependent on the air velocities in the duct and therefore decreasing the airflow rates
will increase the heat gains to the cooled air and the designed supply air temperatures
in the outlet may not be maintained, implying to poorer cooling capacity of supply air.
The warming up effect can be more important when the difference between the
maximum and minimum airflow rates is big and as a result the air velocities decrease
considerably when airflow rates are decreased. This effect should be evaluated and
considered already in the design process when designing the duct system and its
components.
2.5 Energy aspects

One of the main advantages of VAV systems compared to CAV is energy
conservation aspect. Since the VAV system works according to the cooling demand of
the zones compared with systems based on the “sum of the peaks”, considerable
savings in costs for air distribution (fan system), air heating and cooling can be
achieved.
The energy needed for HVAC system can be divided into three main parts:
• Electrical energy
• Heating energy
• Cooling energy
2.5.1 Electrical energy savings

A variable air volume system changes the supply airflow rate according to the need of
the supported room or zone. Generally the average airflow rate in the air handling unit
is somewhat between 50 – 70% of the design air flow rate [33]. Due to the decrease of
the average airflow, the need of electrical energy for air distribution will also decrease.
The power demand needed for a fan (electrical energy) depends on the airflow rate, the
fan total pressure and the total efficiency of the fan and can be expressed as follows
[49]
:
V& ⋅ ∆ ptot
W& t = [W] (2.3)
η tot
Where,
W& t total power input to the fan motor, [W]
V& airflow rate through the fan, [m3/s]
19
∆ptot fan total pressure, [Pa]
ηtot total efficiency of the fan system
The fan total pressure ∆ptot is a function of the airflow rate (see equation 2.4), where
the exponent x in the equation depends on the flow conditions in the system.
∆ ptot = f (V&VAV
x
) (2.4)
For turbulent airflow conditions the system characteristics follow the relationship
where x = 2. For laminar flow conditions the value x = 1. Heating, ventilating and air-
conditioning systems generally work under turbulent flow conditions [4]. However,
when the flow conditions in the system change, as it happens in a VAV system, the
system resistance curve can deviate widely from the relationship where x = 2.
Moreover, as a result of static pressure control in a VAV system, the fan total pressure
is a great extent dependent on the controlled pressure level in the system and the
relationship presented in equation 2.4 can be written:
∆ ptot = f (const + V&VAV

x
) (2.5)
Here the exponent x varies from 1 to 2 and can be mainly dependent on the flow
conditions in the air handling unit, if the static pressure sensor is installed in the
beginning of the duct system. For example in figure 2.8 is presented a VAV system
with the static pressure control after the air-handling unit. The static pressure is kept
on a level that can assure a proper work of VAV terminal units/diffusers at the design
airflow rate. It accounts typically 20 - 40% of the total fan pressure and the size of the
exponent x is mainly dependent on the flow conditions in the air handling-unit.
AHU ∆pSP =∆pduct

Total SP
pressure ∆pAHU
rise, ∆ptot
Fan speed n1 ∆ ptot = f (const + V&VAV
x
)
n2
System curve
∆pdesign n3
n4 Fan curve
∆pAHU
∆pS
∆pduct
V&design Airflow rate, V&
Figure 2.8. Theoretical VAV supply fan and system curves with static pressure
control in the system. AHU- air handling unit; SP- static pressure
sensor; ∆pduct – pressure drops in the duct system with design airflow
rate, which correspond to a set-point in the static pressure control ∆pSP;
∆pAHU- pressure drops in the air handling unit.
20
In variable airflow conditions the total efficiency of a fan system ηtot is also varying,
depending both on the airflow rate and total pressure rise of the fan.
η tot = f (V& , ∆ptot ) (2.6)
The efficiency of the fan itself falls rapidly when the airflow rate gets below 20-30%
of design airflow rate, as shown in figure 2.9. The diagram is calculated for a very
efficient fan [16]. The calculations are done for two different assumed design points on
the fan curve with the nominal speed. The diagrams are made for a quadratic system
curve ∆ p tot = f (V& 2 ) and with no static pressure control in the system. QUARTZ
100
100% Design point 1
80%
80
Fan axis efficiency
Design point 2
60%
60
40%
40
20%
20
0%
0
0% 20% 40% 60% 80% 100%
Fraction of design airflow rate

Figure 2.9. One example of the change in fan axis efficiency with decreasing
airflow rates. The calculations are done for a very efficient fan [16]. The
diagrams are calculated for two different assumed design points and for
quadratic system curve ∆ p tot = f (V& 2 ) . Design point 1 corresponds to
5500 m3/h, 1375 Pa; design point 2 to 4300 m3/h, 1720 Pa.
The drive system efficiency, including motor and drive, starts to fall already at 40-
60% of design airflow rate, as can be seen in figure 2.10. The figure is based on the
calculations done for the motor-drive system with 4.0 kW ABB Motor and Danfoss
VLT 6000 frequency converter. The data for the calculations is obtained from the
measurements done by Åström [68]. The diagrams are calculated for two different
assumptions that the fan power is proportional to either the cube or the square of the
airflow rate, which correspond to turbulent or laminar pressure drops in the system.
QUADRELLA
21
100%
Total drive system efficiency

W& t= f (V& 2 )
80%
60%
W& t= f (V& 3 )
40%
20%
0%
0% 20% 40% 60% 80% 100%
Air flow Fraction of design
rate as fraction of ariflow
design rate
air flow rate
Figure 2.10. One example of the change in motor-drive efficiency with decreasing
airflow rates. The calculations are done for the motor-drive system with
4.0 kW ABB Motor and Danfoss VLT 6000 frequency converter and
are based on the measurements done by Åström [68]. The diagrams are
calculated for two different assumptions that the fan power is
proportional to either the cube or the square of the airflow rate
(turbulent or laminar pressure drops in the system).
According to equations 2.3 and 2.5, the total fan energy consumption can be written as
a function of airflow as follows:
α
W& t = f (const + V&VAV ) (2.7)
In a VAV system with adjustable speed drive control with frequency converter the α
value in the equation 2.7 is mainly dependent on the proportion of duct static pressure
set-point level compared to the total fan pressure. The higher the static pressure
control level is compared to total pressure, the smaller the α value gets and the more
linear is the relationship between the fan power and airflow rate [33, 49]. The value of
the constant in equation 2.7 depends mainly on the controlled pressure level in the
system.
A schematic picture of the energy consumption of a fan in a relation with varying

airflow is given in figure 2.11. With no constant static pressure level kept in the
system and with the ideal airflow control the fan power is theoretically a function of
airflow rate raised to a power of 3. With constant static pressure in the system and
with adjustable speed drive control of airflow, the value for α in equation 2.7 is
between 1 and 3.
22
Fan
power, W& t
W& t design
constant static
pressure in the system
α
W& t = f (const + V&VAV )
α≈2 no constant static
and ideal control case
W& t = f (V&VAV
3
)
Airflow rate, V& V& design
Figure 2.11. Theoretical fan energy consumption dependence on airflow rates. With
no constant static pressure level in the system and with the ideal airflow
control, the fan power is theoretically a function of airflow rate raised to
a power of 3. With constant static pressure in the system and with
adjustable speed drive control of airflow, the value for α is between 1
and 3.
Electrical energy efficiency of a ventilation system is commonly characterized by the

Specific Fan Power (SFP) value. It is calculated by dividing the electrical power
demand of all the fans in the air distribution system by the highest value of supply
airflow rate or exhaust airflow rate through the building under design load conditions
[49]
:
∑ W& t
design
SFP = [kW/(m3/s)] (2.8)
V& design
Where,
SFP Specific Fan Power, [kW/(m3/s)]
W& t Design Design fan power, [m3/s]
V& Design Design airflow rate, [m3/s]
Since the fan electricity varies with the airflow rate, the specific fan power in a VAV
system can also be evaluated for different airflow rates in the system. When
calculating the SFP values for other airflow rates than for the design airflow rate, the
maximum of the supply/exhaust airflow rate should be used in e.q. 2.8.
The variation in SFP value can be written as a function of airflow as follows:
β
SFPVAV = f (const + V&VAV ) (2.9)
Where,
SFPVAV Specific Fan Power for instant conditions in a VAV system, [kW/(m3/s)]
23
The value of β in equation 2.9 is also dependent on the pressure control set values in
the system. The higher the pressure set value is compared to the total pressure, the
more linear dependence is there between the SFPVAV value and airflow rate in the
system.
Figure 2.12 describes the SFPVAV value as a function of airflow with and without
pressure control in the system. With no constant static pressure in the system and with
ideal airflow control of the fans, the SFPVAV is theoretically a function of airflow rate
in square. With constant static pressure in the system and with adjustable speed drive
control of airflow the value β is lower than 2, sometimes even lower than an unity.
Specific Fan
Power, SFP
SFP design
β
SFP VAV = f (const + V&VAV ) Constant static
β<1
No constant static
and ideal control case
SFP = f (V&VAV
2
)
Airflow rate, V& V& design
Figure 2.12. Specific Fan Power dependence on airflow rates. With no constant
static pressure in the system and with the ideal airflow control of fans,
the SFP is theoretically a function of airflow rate raised to a power of 2.
With constant static pressure in the system and with adjustable speed
drive control of airflow, the value β is sometimes less than 1.
It is recommended that the SPF value for a new building project should not exceed 2,0
kW/(m3/s) and for rebuilding and conversion work the maximum SFP 2,5 kW/(m3/s)
should be aimed [7, 47]. The SFP values are usually considered when the overall
efficiency of the air handling system is aimed to be improved.
2.5.2 Thermal energy savings

The power demand for central air heating can be determined as follows:
Q& h = V&VAV ⋅ ρ ⋅ c p ⋅ (t sup ply − t rec ) [W] (2.10)
Where,
Q& h power needed for central heating, [W]
V&VAV instant airflow in a VAV-system, [m3/s]
ρ density of air, [kg/m3]
cp specific heat capacity of air, [J/(kg K)]
tsupply supply air temperature, [°C]
24
trec air temperature after the heat recovery, [°C]
According to equation 2.10, when the airflow rate is decreasing, the amount of heat
needed for supply air heating is decreasing respectively.
Additional energy savings are achieved with the air-to-air heat recovery system
installed in the air handling system. This can be done by returning exhaust air in the
air-handling unit, which is common system type in USA, or by air-to air heat
exchangers.
In Sweden heat is most commonly recovered by air-to-air heat exchangers. The effect
of an air-to-air heat exchanger to an additional energy needed for supply air heating in
a CAV system can be seen on figure 2.13. According to the figure, when decreasing
the supply air temperature the need of heat for supply air heating will decrease
substantially. For example: in the heat recovery system with the temperature efficiency
of ηT = 0,5 the supply air temperature is decreased from +19°C to +15°C. As a result
the need of heating the supply air decreases from 12 MWh/year/(m3/s) to 4,3
MWh/year/(m3/s) respectively, resulting in 7,7 MWh/year/(m3/s) savings in additional
heat energy (see fig. 2.13).
Moreover, the temperature efficiency of the heat recovery increases with decreased
airflow rates, due to the increased ratio of heat exchanger area to airflow rates. This
means that in practice, in a VAV system the need of supply air heating will be
somewhat lower than shown in figure 2.13.
qh
Stockholm
MWh/year ηT=0.50 Temperature-
texh +25°C
m3/s 15 operation 3000 h/year efficiency ηT texh
Exhaust air
10 ηT=0.60
Ambient air
trec tsupply
5 ηT=0.70
Q& h
0
+14 +15 +16 +17 +18 +19 +20
Supply air temperature tsupply,°C
Figure 2.13. The additional energy needed for air heating in a CAV air supply unit
with different heat recovery temperature efficiencies. The duration
diagram responds to the location of Stockholm. Example: the supply air
temperature is decreased from +19°C to +15°C and the heat recovery
system works with the temperature efficiency of ηT = 0,5, additional
heating energy needed for supply air can be decreased from 12
MWh/year/(m3/s) to 4,3 MWh/year/(m3/s).
25
2.5.3 Energy savings for cooling
The cooling capacity needed for central cooling is as follows:
Q& c = V&VAV ⋅ ρ ⋅ ∆h [W] (2.11)
Where,
Q& c power needed for central cooling. [W]
V&VAV instant airflow in the VAV-system, [m3/s]
∆h the enthalpy decrease in the system, [kJ/kg]
In temperate climates extra advantages can be taken from free-cooling, when the
outdoor temperature is equal or lower to the supply air temperature [14]. This also gives
an advantage to VAV system in existing buildings, where central cooling coil inside
the air handling system is not installed. Due to lower airflow rates, lower supply air
temperatures with better cooling capacity can be used if suitable supply air devices are
installed, which should avoid the risk of draught.
26
3 FIELD MEASUREMENTS- TWO CASE
STUDIES
Two different case studies were carried out during the field measurements. The two
office buildings studied were both operating with a VAV system without active
control dampers in the duct system. One of the buildings was an about 40 years old
office building, where the existing CAV ventilation system was replaced with a VAV
system during the summer of 2003. The second case study building consists of two
building parts where one is completely new and the other one fully renovated. This
building was taken into use in April 2004.
The purpose of selecting two case study objects, where one is old and one is new, was
to study if there are any differences in the function of the air conditioning system,
especially from the view of the indoor climate. Both buildings have similar VAV
systems with pressure independent variable supply air diffusers, which supply the air
to the room with constant temperature all year around.
3.1 Case study 1- EDIT-building at Chalmers

University campus
The first case study was carried out in an existing office building (built in early 1960-
ies), where the reconstruction of the air handling system was made three years ago.
Figure 3.1 shows a photo of the building and a typical floor plan. The ventilation
system, that supports 3500 m2 of office area (5 floors), was changed from a constant
air volume system (CAV) to a variable air volume system (VAV). The aim was to
improve indoor climate in rooms (the previous system had no central cooling) and also
achieve energy savings.
Figure 3.1. A photo and a scheme of a typical floor plan in case study 1- EDIT-
building at Chalmers University of Technology in Gothenburg.
27
The office building has totally 107 office rooms on 5 different floors. The air
distribution in the rooms is with mixing ventilation from ceiling diffusers. The exhaust
air from each room is transferred to the corridor through a transfer air grille above the
doors. The rooms are heated by hydraulic radiators below the windows.
The existing approx 15 years old air-handling unit, that has a regenerative air-to-air
heat recovery, and the original duct system were preserved during the renovation. The
old supply air devices were changed to pressure independent variable air volume
diffusers. Consequently no extra active control dampers were installed in the supply
air system. Variable frequency inverters for fan speed control were installed for both
supply and exhaust air fans in order to maintain a specific static pressure in the main
ducts near the air-handling unit. The supply and exhaust airflow rates are balanced on
each floor by measuring the supply and exhaust airflow rates on the main ducts and
controlling the exhaust airflow rate with a damper installed in the main duct next to
the exhaust air shaft. A schematic picture of the installed VAV system in this building
is presented in figure 3.2.
In the figure:
1 Supply air fan with a variable
3 frequency inverter
2
2 Exhaust air fan with a variable
5 frequency inverter
4 3 Pressure sensor
4 Airflow measurement device
1 5 control damper for the exhaust
3
4 air
6
6 Pressure independent variable
supply air diffusers
Figure 3.2. A scheme of the installed VAV system in case study 1- EDIT-building
at Chalmers University of Technology in Gothenburg.
A cooling coil was installed in the air-handling unit during the rebuilding. It is
supplied from the district cooling system of the campus area. The data of the
ventilation system before and after renovation are presented in table 3.1. Due to the
lack of data about the energy consumption with the old HVAC system, the numbers
given in table 3.1 are calculated with the simulation program BV2 [13]. The aim was to
compare and predict the effect of rebuilding the system from CAV to VAV.
The air-handling system is in operation during working hours. After the renovation the
supply air temperature from the central air handling unit is approx. +13 to +14°C all
year around. The duct system was preserved during the renovation and is not
insulated. Therefore the supply air warms up to some extent in the ducts before
reaching the outlets.
28
Table 3.1.
The calculated parameters of the air-handling system before and after the renovation in
the EDIT-building, Chalmers. The temperature efficiency is given as an average over
the year.
System parameter Before rebuilding After rebuilding
CAV system VAV system
Design airflow rate 5,6 m3/s 5,6 m3/s
Minimum sum airflow of supply diffusers - 0,86 m3/s
Operation time ~3500 h/year ~3500 h/year
Supply air temperature 18°C 15°C
Exhaust air temperature 23 °C 23°C
Heat recovery- temperature efficiency average 75% average 78%
Specific fan power- SFP at design airflow 2,5 kW/m3/s 2,5 kW/m3/s
Electrical energy needed for fans 14,8 kWh/year, m2 7,7 kWh/year, m2
Heat energy needed 6,6 kWh/year, m2 0,05 kWh/year, m2
Cooling energy needed for cooling the air - 16,5 kWh/year, m2
Heat energy needed for radiators 83,2 kWh/year, m2 77,8 kWh/year, m2
Outside the working hours the ventilation system is switched off or running under
minimum air flow rates. Night cooling is applied summertime. If the room temperature
increases higher than +21°C during non-working hours and the outdoor temperature at
the same time is below +17°C, all the supply air devices will get a signal for full
opening and the air handling unit starts. When the room temperature gets lower than
+23°C or if the outdoor temperature gets higher than +17°C the air handling unit stops.
All the variable supply air diffusers are equipped with room temperature and presence
sensors. Each supply air device is programmed for two low airflow rates and one
maximum airflow rate. If the room is empty and the room temperature is below
+23°C, the supply air device is working with the minimum airflow rate 5 l/s. When
someone enters the room the airflow rate increases to 10 l/s (which corresponds to
approx 1,1 ACH in a typical room with the size of 11m2). If the room temperature
increases over +23°C, the airflow rate from the diffuser increases up to the maximum
40 l/s (which corresponds to 4,5 ACH in a typical room with the size 11m2) in order to
keep the room temperature on the desired level.
To avoid simultaneous cooling and heating in rooms, a dead band of 2ºC in set
between the set points of supply air devices and the radiators thermostat valves.
3.2 Case study 2- Academicum and

Anatomihöghuset at Sahlgrenska Academy
The second case study in this project was carried out in a modern office building in
Gothenburg. The new administration building for the Sahlgrenska Academy at
Gothenburg University (see figure 3.3) was taken in operation in spring 2004 and
consists of two parts.
29
N
Figure 3.3. A photo and a scheme of the building in case study 2 - Academicum and
Anatomihöghuset at Sahlgrenska Academy in Gothenburg. Building A
marks the new building part - Academicum, building B is the fully
renovated building part - Anatomihöghuset.
The newly built part of the building- Academicum has 1820 m2 of premises area
(LOA), which includes offices (14 office rooms in total), conference rooms, the
Faculty Club and a lecture hall. The total brutto area of the building is 2500 m2 (BTA).
The other, older part of the building Anatomihöghuset consists of mainly office rooms
(65 office rooms in total) with total premises area of 1967 m2 (LOA). The total brutto
area is 2500 m2 BTA. This building part was originally built in the 1960ies, but only
the radiator system was kept during the renovation. All other building services systems
have been fully changed, including the ventilation system.
Floor heating is used in the majority part of the Academicum, whereas office rooms
have hydraulic radiators under the windows. Both cooling and heating systems are
connected to a borehole heat pump/water chiller system.
The ventilation system in the both building parts works with variable air volume
(VAV) operation. Both buildings have their own air-handling unit, which consist of
filters, regenerative air-to-air heat exchanger, a fan system with variable frequency
inverters, heating and cooling coils. Heating coils are installed on the exhaust side of
the system for dumping the heat coming from the water chiller condensers in
summertime. They also can be used for heating the supply air through the regenerative
heat exchanger. The heat exchanger is designed to raise the outside air through to the
required supply air temperature +13°C - +14°C at the design outdoor conditions with
-16°C temperature. The supply air temperature is kept the same all year around. Due to
the heat gains in the duct system the supply air temperature to the rooms is approx
+15°C. The duct system is insulated with 30-40 mm mineral wool insulation in main
and branch ducts. The connection ducts are insulated in the Academicum building part
with 30mm mineral wool. In Anatomihöghuset building the connection ducts are not
insulated.
The VAV system is built up similarly to the system in Case study 1 (see figure 3.2). A
constant static pressure is maintained in the main ducts near the air-handling unit. The
supply and exhaust airflow rates are balanced on each floor by measuring the supply
and exhaust airflow rates on the main ducts and controlling the exhaust airflow rate
with a damper installed in the main duct next to the exhaust air shaft.
30
The air handling system is in operation during working hours. Nighttimes the
ventilation system is switched off or running with minimum airflow rates. The air
distribution to the office rooms is with mixing ventilation from variable supply air
diffusers and the exhaust air is transferred from each room to the corridor through a
transfer air grille above the door. The exhaust air is typically taken from the corridors.
The supply air devices installed in the office rooms are similar to the devices in the
case study 1. They are pressure independent variable supply air diffusers and therefore
no extra active control dampers were needed in the supply air duct system.
Each variable supply air diffuser is programmed for two low airflow rates and one
maximum airflow rate. If the room is empty and the room temperature is under +23°C,
the diffuser is working with the minimum airflow rate 7 l/s (which corresponds to
approx 0,7 ACH in a typical office room). When someone enters the room the airflow
rate increases to 10 l/s (which corresponds to approx 1 ACH in a typical room with the
size of 13 m2). If the room temperature increases over +23°C, the airflow rate
increases up to maximum 30 l/s (which corresponds to approx 3 ACH in a typical
office room). In a few conference rooms the maximum airflow rate per device is 50
l/s.
The lecture hall in the Academicum building part has a demand-controlled ventilation
(DCV) with a CO2 sensor with the set point of 700 ppm. The supply air is delivered to
the room from grilles located under the chairs with the supply air temperature of
+19°C. The supply air is warmed up from +15°C, which is supplied from central
AHU, via a re-heating coil.
The design parameters of the ventilation systems in both building parts are presented
in table 3.2.
Table 3.2.
The design parameters of air-handling systems in case study 2: Academicum and
Anatomihöghuset at Sahlgrenska Academy.
System parameter Academicum Anatomihöghuset
Design airflow rate 5,0 m3/s 3,6 m3/s
Maximum total airflow of supply diffusers 4,2 m3/s 3,0 m3/s
Minimum total airflow of supply diffusers 0,71 m3/s 0,66 m3/s
Operation hours 3500 h/year 3500 h/year
Supply air temperature to the rooms 15°C 15°C
Exhaust air temperature 22°C 22°C
Heat recovery- temperature efficiency min 82% min 82%
Specific fan power- SFP at design airflow 2,1 kW/m3/s 1,9 kW/m3/s
31
3.3 Conducted measurements in the case studies
3.3.1 Indoor climate measurements
Thermal comfort and noise measurements were carried out in a number of randomly
selected office rooms in both case studies. The thermal comfort measurements were
mainly focused on the risk of local thermal discomfort in rooms. The measurements
included room temperature and air velocity measurements in order to evaluate the risk
of draught.
The noise measurements included measuring the sound pressure levels in the rooms.
Problems with background noise occurred during the measurement time. To evaluate
the noise levels generated by the VAV diffuser itself, it was presumed that it cannot
exceed the minimum measured sound pressure level in the room.
The rooms chosen for the measurements were typical office rooms. Figure 3.4 shows a
typical room and the location of the measuring instrument. The room size and layout
was similar in both case studies. The measuring sensors/transducers were placed at the
level of a sitting person’s head (1,1m above the floor) near the working place (See fig.
3.4).
troom, va, SD
VVAV,
tsupply
Figure 3.4. Typical room and the location of the measuring instrument (the location
marked X) in field measurements. The size of a typical room was ~11 m2
in case study 1 and ~13 m2 in case study 2.
The thermal comfort and noise measurements in every selected room were made at the
conditions of the maximum supply airflow rate to the room and with the supply air
temperature about +15oC, since these are the conditions that may cause a risk of
draught and extra noise [10, 14]. The measurement time was approx 10 minutes. For the
draught measurements the last 3 minutes were used for the calculations.
The information about the instruments used for thermal comfort and noise
measurements are given in APPENDIX 1.
The airflows from the diffusers were measured with the equipment installed into the
device. The control and regulating equipment, as well as the sensors, are built into the
variable supply air diffuser and the simultaneous airflow rate values were read with a
32
palm computer. The airflow rates to the room were controlled with setting the needed
airflow rate through logging into the device with the palm computer.
The supply air temperature and pressure in the duct just before the supply air device
was also measured in the selected rooms.
3.3.2 Energy consumption measurements

The energy consumption of the air-handling systems and total airflow rates of the
systems were monitored during somewhat more than a year. In the first case study
(EDIT-building at Chalmers) the presented results cover the period of January 2004 to
December 2004; in the second case study (Academicum and Anatomihöghuset) the
results are taken from the period May 2004 to April 2005. A small gap in the
measurement period occurred during the second case study in Anatomihöghuset and as
a result there was no values recorded during the period of 5-19 of August.
During the energy consumption measurements following parameters were monitored:

• Outdoor temperature (in the air intake duct before the air handling unit)
• Supply air temperatures after the fan and the heat exchanger
• Supply air temperature after the cooling coil (the supply air temperature to the
building)
• Supply air temperature after the heating coil (case study 1)
• Exhaust air temperature from the building
• Exhaust air temperatures before and after the heating coil (case study 2)
• Exhaust air temperature after the air handling unit
• Supply and return water temperatures of the heating coil (case study 1)
• Supply and exhaust airflow rates with measuring pressure difference with
sensors inside the fan casing (case study 2) or with static pressure difference
measurements over the heat exchanger (case study 1)
• Electric power for supply and exhaust air fans
The sampling time was 60 seconds and the results are presented in the diagrams in
chapter 5 as one-hour average values.
The energy consumption and airflow rate measurements in both case studies were
carried out by CIT Energy Management AB [34, 35], on commission by the building
owner, Academiska Hus. Therefore the measurement techniques and instrumentation
will be not discussed deeply in this thesis work. Only the results for comparison and
energy consumption evaluation will be given in the final results.
3.4 Indoor climate evaluation - questionnaire

Besides physical measurements for thermal comfort evaluation in office rooms,
questionnaires of the users’ perceptions and their preferences were carried out in both
case studies. The use of questionnaires for thermal environment assessment has been
applied in practice in many previous studies [6, 40, 51, 54, 56, 59].
33
The used questionnaire, presented in Appendix C, is based on ISO10551 [31] standard
and consists of questions about indoor environmental parameters such as perceived
room temperature, air movement, air humidity, noise, light and air quality. This kind
of questionnaire was chosen because it has been commonly accepted and used in many
previous studies [56, 58, 59].
Very similar questionnaire as the one accounted here has been carried in hospital
wards [56]. However, instead of asking people to evaluate their indoor environment at a
certain moment, as it was the case in the study in hospital wards, building occupants
were giving their opinion over certain period of the time. Two time periods were
evaluated for both case studies: summer and winter period.
A seven point judgement scale was used for evaluating the perception of air
temperature, air velocity, air humidity, noise, lightning and daylight. The values 1 and
7 corresponded to extreme situations and 4 was assigned as neutral, which can be
considered as an ideal case.
The air quality and perceptions for overall indoor environment were judged on a scale
1 to 7, with the ideal point of 7.
A seven-point scale was also used for assessment of the preference of the parameters
described above. In these questions 1 and 7 corresponded to extremes and 4 described
the situation as “It is good as it is”.
An example of the questions asked for evaluating one environmental parameter is

given in table 3.3.
Table 3.3.
An example of the questions in the questionnaire
1a. How do you feel about the room temperature during this period of the year?
Cold Cool Slightly Neutral Slightly Warm Hot
cool warm
1 2 3 4 5 6 7
1b. Do you find it …?

Extremely Very Uncomfortable Slightly Comfortable
uncomfortable uncomfortable uncomfortable
1 2 3 4 5
1c. Please mark, how would you like to have the temperature during this period of
the year?
Much Colder Slightly It is good Slightly Warmer Much
colder colder as it is warmer warmer
1 2 3 4 5 6 7
The first questionnaire study was carried through in connection with the first case
study, the EDIT-building. The total number of questionnaires distributed in this case
study 104 for evaluating the winter period and 114 for the summer period. The
34
percentage of questionnaires returned was approx 42% both times. This low
percentage was due to the fact that many university workers in that building were
lacking of time to fill it in and not all the distributed questionnaires were returned.
However, the results were obtained from at least half of the office rooms per floor.
The second questionnaire study was conducted in the buildings in Case study 2
(Academicum and Anatomihöghus). In this study 55 questionnaires were distributed
for the summer period and 70 for the winter period. All together 53% of the summer
questionnaires and 67% of the winter ones were answered.
As a result mean values of the occupants´ votes for different parameters were
calculated for both case study buildings for summer and winter period.
35
36
4 LABORATORY MEASUREMENTS
4.1 The test room
Laboratory measurements were carried out in a simulated environment: in a full size
office cube built inside the laboratory hall. A schematic picture of the test set-up is
shown in figure 4.1.
The test set-up consisted of a test room, a supply air fan with frequency inverter and
with a pressure control, a sound attenuator, a supply air heater with an air temperature
control, a flow meter, a pressure independent variable supply air diffuser and
temperature sensors. The temperature sensors were used for monitoring temperatures
in the duct and inside and outside the test room. All sensors were connected via a
logger to a personal computer for monitoring purpose.
CC1
T HC T PS1
F1 SA
TS1 FM
SF TS2
PS2
CC2 TS3
In the figure:
SD
SD Supply air diffuser
SA Sound attenuator PC
Test room
SF Supply air fan
CC1 Frequency inverter with PID control V& = 10-50 l/s
F1 Duct filter
FM Flow measurement
HC Heating coil TS4
CC2 Control center
TS Temperature sensor
PS1 Pressure sensor static pressure control
PS2 Pressure sensor
T Thermometer
Figure 4.1. A scheme of the test set-up for the laboratory experiments.
The test room was constructed with plaster boards on a wooden framework. The
internal dimensions of the test room were 3,9 x 2,8 x 2,7m. To imitate a common
office environment the room was filled with usual office equipment: a table, a chair, a
computer and lighting.
The internal heat loads were simulated with a computer (120W), a dummy (80W) and
lighting (total 370W). The artificial light consisted of four fluorescent tube fittings (85
W heat gains per one tube) plus a table lamp (30W heat gains).
Photos of the test room are presented in figure 4.2.
37
Figure 4.2. The test room in the laboratory measurements. The room was
constructed to simulate the conditions of a common cell office room. A
placement of heat loads: computer, dummy and lighting can be seen on
the right photo.
The air was supplied from a variable supply air diffuser mounted in the middle of the
ceiling. Two different arrangements were tested: one with the diffuser free from the
ceiling and one with the diffuser in the suspended ceiling. Without suspended ceiling
the height from the ceiling to the discharge area of the device was 30 cm and from the
device to the floor 2,4 m. With suspended ceiling the latter height was increased to
2,7m (the device itself with the duct was installed above the test room box to simulate
the suspended ceiling case).
For the exhaust air a transfer air device was installed above the door, under the ceiling.
A schematic picture of the test room layout is given in figure 4.3.
In the figure:
1 supply air diffuser D160

2 transfer air device D160
3 dummy
4 personal computer
2,8 m
5 table lamp,
6 ceiling lamp
7 ceiling lamp
8 window
3,9 m
Figure 4.3. The layout of the test room. The size of the test room was ~11 m2.
The airflow rate to the room varied between 10-50 l/s, which corresponds to 1,2 – 6,1
ACH. The supply air temperature was kept constant +15°C and for all the tests the
operative temperature in the occupied zone and also outside the office cube was kept
22 ± 1°C. This corresponds to the A level comfort class according to the design criteria
38
for office rooms [11]. The classification was described in more detail in chapter 2.1.1.
The temperatures inside and outside the room were monitored constantly.
A variable frequency inverter for fan speed control was installed for the supply air fan
in order to maintain a specific static pressure in the duct. The set value for constant
static pressure in the duct was approx. 50 Pa.
Testing was made for different airflow rates. The heat loads were adapted to the
airflow in order to obtain the correct room temperature. The testing was performed
after steady-state conditions in the room were fully established.
A schematic picture of the variable VAV diffuser tested in the laboratory is shown in
figure 4.4. The work principle of the diffuser is following. The distance between the
lamella plates (se figure 4.4) varies according to the required airflow rate and the set
static pressure before the device and is controlled by a traversing motor, which gets
impulses from the controlling sensor. The control and regulating equipment as well as
the sensors are built into the supply air device and the simultaneous values can be read
with the computer.
Supply duct
connection
Traversing motor
V (l/s), tsupply , vd (m/s)
room temperature Lamella plates

sensor
Electronics and
controls
Figure 4.4 Tested variable supply air diffuser
4.2 Conducted measurements in the laboratory

Thermal comfort measurements were carried out with constant supply air temperature
+15°C, but with different airflow rates. The parameters used in the experiments are
given in table 4.1. There is a difference between heat load and cooling power values in
the table. This is due to the heat transmission through the test room walls.
Table 4.1.
The test cases completed in thermal comfort measurements
Case Mounting Air flow tsupply Cooling Balancing Nr. of measuring
condition rate, l/s °C capacity, W heat load, W points
1 no SC* 10 15 85 170 27
2 no SC 25 15 210 250 27
3 no SC 50 15 420 570 27
4 SC 10 15 85 170 27
5 SC 25 15 210 250 27
6 SC 50 15 420 570 27
* SC- suspended ceiling
39
Air temperature and air velocities were measured in a number of room points as shown
in figure 4.5. At each position the measurements were taken at 3 heights - 0,1m, 0,6m,
1,1m, which is based on the position of a sitting person [5] (see fig. 4.5). The
measurement period for each measurement was 3 minutes.
The steady-state conditions were monitored using temperature data from sensors
installed inside and outside the room. The operative temperature in the room was
estimated in the position of a sitting person next to the work place, on the level of 1,1
m from the floor. For estimating the operative temperature a plane radiant temperature
for six different directions was measured and the operative temperature was calculated
according to equations 4.1 and 4.2 [9, 30].
t op = a ⋅ t a + (1 − a ) ⋅ t r [°C] (4.1)
0,18 ⋅ (t pr [up ] + t pr [down]) + 0,22 ⋅ (t pr [right ] + t pr [left ]) + 0,30 ⋅ (t pr [ front ] + t pr [back ])

tr =
2 ⋅ (0,18 + 0,22 + 0,30 )
(4.2)
Where,
top operative temperature, [°C]
a constant. For the values of relative air velocity var < 0,2 m/s the value a = 0,5
tr mean radiant temperature, [°C]. The equation 4.2 is used for calculating
mean radiant temperature for a sitting person.
tpr plane radiant temperature, [°C]. The direction given in the brackets in
equation 4.2 is the direction where the temperature is measured.
Every measurement case described in table 4.1 was done in three replicates and the
results are presented as an average over these three measurements.
0,5 m 1,45 m 1,45 m
1 2 3
0,9 m
4 5 6
0,9 m
0,5 m
7 8 9
Figure 4.5. The measurement points in the room and a photo of the measurement
set up.
40
In order to measure the supply airflow rate in the duct, a pressure difference over an
airflow measurement device was measured and the airflow rate calculated
respectively.
The supply air temperature was kept constant +15˚C during the measurement time and
monitored with the temperature sensor installed in the duct.
The thermal comfort in the test room was evaluated with the draught rate (DR) model,
described in chapter 2.1.2. The draught rating was calculated for every measured
point. The reason for focusing on draught is that draught has been considered to be
one of the most common complaints indoors [24] and has been also associated with
VAV air distribution [10, 42, 55].
41
42
5. RESULTS AND DISCUSSION
5.1 Results from the case studies
5.1.1 Thermal comfort in measured rooms
Case study 1
Table 5.1 below shows measured thermal comfort parameters in different rooms in the
Case study 1. The values are given as an average over the measurement period
(approx. 10 minutes). Average air velocity and standard deviation of air velocity are
given as an average over the 3 minutes before the last measured minute (to exclude the
disturbing effect to air movement from people).
Table 5.1.
Measured thermal comfort parameters in selected rooms in Case study 1 with the
maximum airflow rate 40 l/s from the diffuser.
Room Supply air Room Air Standard
nr. temp. temp. velocity, deviation of air
tsupply, °C troom, °C va, m/s velocity m/s, SD
1 14,7 22,7 0,12 0,04
2 15,0 23,4 0,08 0,03
3 14,7 21,4 0,07 0,02
4 15,6 22,6 0,10 0,03
5 15,2 23,5 0,10 0,03
Figure 5.1 below expresses the calculated draught rating in measured rooms. The
draught rates are calculated according to equation 2.1 (given in chapter 2.1.2).
Draught rating DR, %
25
20
>15%- unacceptable level
15
10
0
1 2 3 4 5
Measured room nr
Figure 5.1 Calculated draught rating in measured rooms in Case study 1.

According to thermal comfort standard ISO7730 [30], the draught rating
should be kept lower than 15%.
43
Case study 2
The results from the thermal comfort measurements conducted in Case study 2 are
presented in table 5.2. Here all the results are presented as an average over a 3-minute
period.
Table 5.2.
Measured thermal comfort parameters in selected rooms in Case study 2 with the
maximum airflow rate 30 l/s from the diffuser.
Room Supply air Room Air Standard
nr. temp. temp. velocity, deviation of air
tsupply, °C troom, °C va, m/s velocity m/s, SD
1 15,4 21,3 0,08 0,02
2 15,1 21,3 0,12 0,03
3 15,1 20,3 0,10 0,03
4 14,5 20,7 0,07 0,02
5 14,5 21,0 0,11 0,04
Figure 5.2 below expresses the calculated draught rating in measured rooms. The
draught rates are calculated according to equation 2.1 (given in chapter 2.1.2).
25
Draught rating DR, %
20
>15%- unacceptable level
15
10
0
1 2 3 4 5
Measured room nr
Figure 5.2 Calculated draught rating in measured rooms in Case study 2.

According to thermal comfort standard ISO7730 [30], the draught rating
should be kept lower than 15%.
In thermal comfort standard ISO 7730[30] the elimination of draft is addressed by

placing limits on the allowable mean air velocity as a function of air temperature and
turbulence intensity. The standard recommends limiting the draught rating to 15%,
which would restrict the mean air velocity to 0.12 m/s at 20°C and 0.2 m/s at 26°C.
This applies for a turbulence intensity of 40%, which is typical for indoor office
environments. The measurement results in both case studies show that the mean air
velocities and DR values did not exceed the recommended levels in the conditions
when supply airflow rate was in maximum and supply temperature approx +15°C.
44
5.1.2 Noise levels in measured rooms
Problems with background noise occurred during the measurement. To evaluate the
sound effects generated by the VAV diffuser itself, it was presumed that it cannot
exceed the minimum measured sound pressure level in the room. That means that the
actual noise level generated by the VAV diffuser can be even lower than the measured
value.
In Case study 1 the sound measurements were made in two rooms in the lower floors,
far from the air handling unit, which is locating on the highest floor of the building. In
both rooms the minimum sound pressure level was between 26 dB(A) and 27 dB(A)
with the pressure drop of 107 Pa and with the maximum airflow rate 40 l/s.
In Case study 2 the measurement was done in one office room locating on the upper
floor of the Academicum building (the air handling unit is locating in the basement).
The results yielded the minimum sound pressure to be approx 19 dB(A) with the
pressure drop 98 Pa and with the maximum airflow rate 30 l/s.
The manufacturer has made own laboratory measurements of the sound levels
produced by their VAV device. Figure 5.3 shows noise levels with different airflow
rates and pressure drops measured in the laboratory compared to the measured in the
rooms in operation.
1000
500
300
Pa
200
drop,Pa
40 dB(A)
Pressuredrop,
100
Academicum 35 dB(A)
Academicum
30 dB(A)
Pressure
50 EDIT-building
EDIT-building 25 dB(A)
20 dB(A)
10
5
1 3 5 10 30 50 70
Airflow rate, l/s
Figure 5.3. Results from the noise measurements with different airflow rates and
pressure drops measured on the field and presented together with data
from the device producing company.
45
5.1.3 Indoor climate evaluation - questionnaire
Case study 1
The results from the questionnaire from Case study 1 (EDIT-building, Chalmers) show
that the majority of the people felt their indoor climate to be close to neutral (see
figure 5.4). According to previous research [58, 59], gender does not appear to affect the
perception of thermal comfort. Therefore the diagram accounts all the results together
without separating the votes from men and women.
For evaluated temperature, air velocity and air humidity the values at ±0,5 from the
middle value 4 correspond to 10% of predicted percentage of dissatisfied (PPD) based
on the PPD index model proposed by Fanger [22]. The PPD index establishes a
quantitative prediction of the number of thermally dissatisfied people.
Scale
7 summer 7 hot
very high
winter very strong
6 6 warm
high
value
strong
5 5 slightly warm
vote
slightly high
mean
slightly strong
4 4 neutral
Mean
Voted
3 3 slightly cool
slightly low
Air humidity
Air speed
Noise
Daylight
Temperature
Lightning
slightly weak
2 2 cool
low
weak
1 1 cold
very low
Parameter very weak
Figure 5.4 The results from the questionnaire in Case study 1. The diagram shows
the difference of mean vote from the ideal value “4”, which corresponds
to neutral vote for the given parameter. The error bars represent 95%
confidence limits of the mean value.
As it can be seen from the figure 5.4, the perceived air temperature was evaluated to
be between slightly cool and neutral during wintertime and between neutral and
slightly warm during summer period. Colder sensation of room temperatures during
the winter period was reported mainly from the room occupants on the first floor of
the building, where the cold floor may have affected the results. The higher perception
of the room temperature during summer time was reported to be mainly from the east
side of the building.
In order to analyse if the mean values of different parameters really differ between the
winter and summer cases, a statistical analysis of the results was carried out.
Significance tests were carried out by using Student’s T-test (statistical hypothesis
test), which compares the actual difference between two means in relation to the
variation in the data carried out. A null hypothesis in the T-test is that the means of
two normally distributed populations are equal.
46
The difference of two mean values is statistically significant, meaning that the
difference in the values is not occurring by a chance alone, if the tested probability of
this occurrence is more than the selected significance level. The chosen significance
level in the analysis accounted here is 95%, meaning that the conclusions have a 5%
chance of not being true (p = 0,05). For example, if the calculated T value exceeds the
tabulated value given in the T-table for p = 0.05, then there is a 95% chance of the
means being significantly different.
According to the results of the two-sample two-tailed T-test, the statistical significance
has been indicated for the difference between the mean values of summer and winter
occasion for air temperature, air humidity, lighting and daylight. The results of the
two-sample two-tailed T-test are presented in table 5.3.
Table 5.3.
The results of the statistical significance test for evaluating the differences of the mean
values of parameters of winter and summer occasion in Case study 1. A two-sample
two-tailed t-test was conducted. In order for the difference between the means to be
statistically significant: |Tcal|>Tcrit
Parameter Significance Degrees of Calculated T Critical T
level freedom value, Tcal value, Tcrit
Air temperature 95% 90 -3,28 1,98
Air humidity 95% 90 -2,10 1,98
Lighting 95% 91 -2,09 1,98
Daylight 95% 90 -3,39 1,98
In general no sensation of draught was indicated from the subjective data gathered
from the building occupants in Case study 1. Moreover, the air movement was
reported to be slightly slow. This can be directly connected to the perception of higher
temperatures, where higher air velocities are accepted in order to compensate the high
air temperature. This also revealed from the results from the questions about the
preferences for different parameters. As can be seen from the figure 5.5, during the
warm period of the year the building occupants would have preferred slightly higher
air movement, but slightly colder air temperature. For the winter occasion this
preference was evaluated to be close value 4, which states the condition to be “as good
as it is”.
The perception of dry air during winter period is quite common in office buildings due
to the supply of relatively dry outdoor air by ventilation systems.
The perception of slightly strong day light during summer time can be related to the
use of insufficient sun protection. Even if the questioned people did not have strong
complaints about noise levels, the majority of the people would have preferred the
sound level to be slightly lower (see fig. 5.5). However this may be related to the
background noise that was indicated during the noise measurements. Noise from
outside, people talking in the corridors and noise from the elevators can be disturbing
for the people working in the office rooms.
47
Scale
7 summer 7 much warmer
much higher
winter much stronger
6
6 warmer
higher
stronger
5 slightly warmer
5
Preference
slightly higher
slightly stronger
4
4 it is good as it is
3 3 slightly colder
slightly lower
Air humidity
Air speed
Noise
Daylight
Temperature
Lightning
slightly weaker
2 2 colder
lower
weaker
1 1 much colder
much lower
much weaker
Parameter
Figure 5.5 The preference of the evaluated parameters in Case study 1. The ideal
value “4” corresponds to “it is good as it is” for the given parameter.
The error bars represent 95% confidence interval of the mean value.
The perceived air quality and the average of the votes given for the indoor
environment are presented in figure 5.6. As can be seen from the figure, both of the
parameters have been evaluated to be close to acceptable.
Very
good 7
summer Scale
winter 7 very good
6
6
votevalue
good
5 5 slightly good
mean
4 4 neighter bad nor good

Mean
Voted
3 3 slightly bad
2 2 bad
Very
1 1 very bad
bad
environment
Air quality
Indoor
Parameter
Figure 5.6 The results from the questionnaire in Case study 1. The figure shows
the mean vote for perceived air quality and indoor environment. The
“7” corresponds to the ideal case and the vote “4” is acceptable. The
error bars represent 95% confidence interval of the mean value.
48
Case study 2
Figure 5.7 shows the differences of mean values from the ideal case for evaluated
indoor climate parameters in Case study 2 (Academicum and Anatomihöghuset,
Sahlgrenska Academy). According to the figure, the majority of the people evaluated
their indoor climate to be close to neutral, which states the ideal condition when
looking at the 7-point scale on the diagram.
Scale
7 hot
7 summer very high
very strong
winter
6 6 warm
high
strong
votevalue
5 5 slightly warm
slightly high
slightly strong
mean
4 4 neutral
Mean
Voted
3 slightly cool
3 slightly low
slightly weak
Air humidity
Air speed
Noise
Daylight
Lightning
Temperature
2 cool
2 low
weak
1 cold
1 very low
very weak
Parameter
Figure 5.7 The results from the questionnaire in Case study 2. The diagram shows
the difference of mean vote from the ideal value “4”, which corresponds
to neutral vote for the given parameter. The error bars represent a 95%
confidence interval of the mean value.
According to the results of two-sample two-tailed T-test there is a statistically

significant difference between the summer and winter conditions for air temperature,
air velocity, air humidity, lighting and daylight. The results of the two-sample two-
tailed t-test are presented in table 5.4.
Table 5.4.
The results of the statistical significance test for evaluating the differences of the mean
values of parameters of winter and summer occasion in Case study 2. A two-sample
two-tailed t-test was conducted. In order for the difference between the means to be
statistically significant: |Tcal|>Tcrit
Parameter Significance Degrees of Calculated T Critical T
level freedom value, Tcal value, Tcrit
Air temperature 95% 73 -4,23 1,98
Air velocity 95% 73 3,69 1,98
Air humidity 95% 72 -2,79 1,98
Lighting 95% 71 -2,22 1,98
Daylight 95% 72 -2,53 1,98
The winter occasion was evaluated to be slightly cold regarding room temperature.
However, the summer condition was evaluated to be slightly warm. The sensation of
slightly low air movement during the summertime can be related to the sensation of
49
higher room air temperature as described before. This can be also seen from the
evaluation of the preference, shown in figure 5.8. During the warm period of the year
the building occupants would have preferred slightly higher air movement, but slightly
colder air temperature.
The air humidity was perceived to be slightly low, which is common complaint during
winter period. The results also show similar perception for summer period.
Scale
7 summer 7 much warmer
much higher
winter much stronger
6 6 warmer
higher
stronger
5 5 slightly warmer
Preference
slightly higher
slightly stronger
4
4 it is good as it is
3 3 slightly colder
slightly lower
Air humidity
Air speed
Noise
Daylight
Temperature
Lightning slightly weaker

2 2 colder
lower
weaker
1 1 much colder
much lower
much weaker
Parameter
Figure 5.8 Preference of the evaluated parameter in Case study 2. The ideal value
“4” corresponds to “it is good as it is” for the given parameter. The
The perceived air quality was evaluated as close to acceptable, as can be seen in figure
5.9. Here the dust and stuffy air were reported to be affecting the air quality. The
indoor environment was evaluated to be not good but also not bad.
In general, the questionnaires carried out in both case studies revealed no sensation of
draught in rooms. Also the noise levels in the rooms were perceived to be acceptable.
These parameters can have the highest risk of occurring in a VAV system [10, 14, 42, 55].
The most common complaint was slightly high room temperature during summer time,
and slightly cold sensation during wintertime. The cause for this can be affected by
many things and more detailed monitoring about air temperatures in those rooms,
where the complaints occurred the most, should be carried out.
50
Scale
summer 7 very good
7
winter
6 6 good
value
meanvote
5 5 slightly good
VotedMean
4 4 neighter bad nor good
3 3 slightly bad
2 2 bad
1 1 very bad
Air quality
environment
Indoor
Parameter
Figure 5.9 The results from the questionnaire in Case study 2. The figure shows
the mean vote for perceived air quality and indoor environment. The
5.1.4 Energy consumption monitoring
Case study 1
The energy consumption of the air-handling units was monitored in both case studies
during somewhat more than one year of measurement period. Figure 5.10 presents the
monthly energy consumption measured in Case study 1 (EDIT-building, Chalmers).
The energy needed for cooling and after heating is calculated from measured airflow
rates and air temperatures inside the air-handling unit. The inaccuracy of the results for
heating and cooling energy is ±10%.
As figure 5.10 shows, there was no need for heating the supply air with the heating
coil. The annual consumption of the electrical energy for air distribution was 23,6
MWh/year, which corresponds to 6,7 kWh/year per m2 of the brutto area (BTA). The
calculated assumption for the building in the same case study was 7,7 kWh/year per
m2 BTA (see table 3.1 in chapter 3.1). According to previous study, that was done in
123 office buildings in Sweden, the average electrical energy used for fans in an office
building is 17,9 kWh/year per m2 [17]. The study was done in office buildings with
different type of ventilation systems.
The cooling capacity needed for supply air during this one year period was 22,5
MWh/year, which corresponds to 6,4 kWh/year per m2 BTA . The calculated
assumption for the cooling energy was 16,5 kWh/year per m2 BTA.
51
Total energy consumption
Heating MWh/yr kWh/yr/m2

Cooling heating
heating coil coil0 0
14 cooling
cooling 22,5
coil coil 6,4
12 return
return fan fan11,8 3,4
Energy [MWh/month]
10 supply
supply fan fan11,8 3,4
8
6
4
2
0
-2
-4
November
June
January
July
September
October
December
March
April
August
February
-6 May
Fans
Month
Figure 5.10. Measured monthly energy consumption in Case study 1. The diagram
presents the energy needed for supply and return air fans, heating and
cooling coils during the period of January to December 2004.
In general, the measured results of the energy consumption in Case study 1 were
somewhat lower compared to the predicted numbers calculated with the reference
building. Moreover, as can be seen from figure 5.11, the average outdoor temperature
in August 2005 [44] was many degrees warmer compared to average outdoor
temperatures for the “normal” year 1961-90. This also explains much higher need for
cooling in August compared to June and July and implies that during the average year
the numbers presented in figure 5.10 can be even lower.
20 2004
1961-90
(monthly average) [ºC]
Outdoor temperature
15
10
-5
November
June
January
July
September
October
December
March
April
August
February
May
month
Figure 5.11. Monthly average outdoor temperatures in Gothenburg during the
measurement period January- December 2004 [44].
52
Case study 2
Figure 5.12 presents the monthly energy consumption for fan energy, cooling capacity
used for cooling coil and heat energy used for heating coil in Case study 2A
(Academicum, Sahlgrenska) during one year of measurement period, from May 2004-
April 2005. The energy needed for cooling and heating of the air is calculated from
measured airflow rates and air temperatures inside the air-handling unit.
According to the measurement results there was no need for thermal energy for
heating the supply air in the system. Also the need for electrical energy for fans was
relatively low. Higher electricity consumption for the exhaust air fan during August
2004 can be related to the water chiller function, where the extra heat dumped by the
condensers in the water chiller system is transferred away by the exhaust air fan,
which then works with maximum speed.
The annual consumption of electrical energy for air distribution was 17,7 MWh/year,
which corresponds to 7,1 kWh/year per m2 BTA. The cooling capacity needed for
supply air during this one year of measurement period was 27,8 MWh/year, which
corresponds to 11,1 kWh/year per m2 BTA.
Heating
Total energy consumption
Cooling
heating coil
20 MWh/yr kWh/yr/m2
cooling coil
heating coil 0 0
return fan
cooling coil 27.8 11.1
Energy [MWh/month]
16
supply fan
return fan 11.2 4.5
12 supply fan 6.4 2.6
-4
November
June
July
January
September
October
December
March
April
August
May
February
Fans
Month
Figure 5.12. Measured monthly energy consumption in Case study 2A. The diagram
presents the energy needed for supply and exhaust air fans, heating and
cooling coils during the period of May 2004 to April 2005. The
inaccuracy of the results for heat and cooling energy is ±10%.
Figure 5.13 gives the data about the outdoor temperatures during the measurement
period of May 2004 to April 2005. As the average outdoor temperature in August
2005 was many degrees warmer compared to the “normal” year 1961-90, the need for
cooling during that month was also higher. This means that during the average year
the numbers presented in figure 5.12 can be even lower.
53
20 2004/2005
1961-90
(monthly average) [ºC]

Outdoor temperature
15
10
-5
November
May
July
January
June
September
December
March
April
August
February
October
month
Figure 5.13. Monthly average outdoor temperatures in Gothenburg during the
measurement period of May 2004 to April 2005 in Case study 2 [44].
Figure 5.14 presents the monthly energy consumption for the fan energy, cooling
capacity used for cooling coil and heat energy used for heating coil in the second
building part of Case study 2B (Anatomihöghuset, Sahlgrenska). The measurements
were done during the same measurement period as in Case study 2A (from May 2004-
April 2005). The energy needed for cooling and heating of the air is calculated from
measured airflow rates and air temperatures inside the air-handling unit.
According to figure 5.14, also this air handling system did not require any thermal
energy for heating the supply air and also the need for electrical energy for fans was
relatively low. The annual consumption of electrical energy for air distribution was 7,7
MWh/year, which corresponds to 3,1 kWh/year per m2 BTA. The cooling capacity
needed for supply air during the one-year period was 6,9 MWh/year, which
corresponds to 2,8 kWh/year per m2 BTA. The cooling energy needed during summer
period was very low compared to the results from Case study 2A building. However,
since there was a problem with measuring equipment between 5-19th of August and no
values were recorded during this period, the results in reality may have been somewhat
higher than revealed in figure 5.14.
54
He ating Total energy consumption
Cooling heating coil
MWh/yr kWh/yr/m2
5 cooling coil
heating coil 0 0
return fan
4 cooling coil 6.9 2.8
Energy [MWh/month] supply fan
return fan 5.1 2.0
3
supply fan 2.6 1.1
2
-1
November
June
July
January
September
October
December
March
April
August
May
February
-2
Fans
Month
Figure 5.14. Measured monthly energy consumption in Case study 2B. The diagram
presents the energy needed for supply- and return air fans, heating and
cooling coils during the period of May 2004 to April 2005. The
inaccuracy of the results for heat and cooling energy is ±10%.
5.1.5 Airflow rate and fan power
Case study 1
Figure 5.15 gives an overview of the measured supply airflow rates a relation to the
outdoor temperatures in the building in Case study 1 (EDIT-building, Chalmers).
Theoretically the airflow rates should decrease when outdoor temperature is

decreasing. This effect can be seen in the figure. However, according to the
measurements, the average airflow rate was approx 2/3 of the maximum also during
the cold period of the year, which implies that the heat loads in the rooms were
relatively high all year around.
Figure 5.16 gives a duration diagram of the total supply airflow rate presented together
with corresponding energy consumption of the supply air fan. According to the
diagram, both of the durations seem to have a good correlation: with the decreased
airflow rates the energy effect for the fans decreases respectively. Total operating
hours of the system was approx 4400 hours/year.
The design airflow rate for the air-handling unit in Case study 1 is 5,6 m3/s. However,
this value was exceeded during some hours during the measurement period and it
occurred mainly during the warm period of the year (see fig. 5.15). The maximum
measured airflow rate was 6,3 m3/s and maximum measured electricity consumption
was 7,1 kW.
55
7
Supply airflow rate [m³/s]

6
0
-10 -5 0 5 10 15 20 25 30
Outdoor temperature [ºC]
Figure 5.15. Supply airflow rate as a function of outdoor temperature in Case study
1. The design airflow rate for the air handling unit is 5,6 m3.
7 Airflow rate
Supply m3/s
airflow
rate [m³/s]
Electricity
Electric kW
Power
Airflow
3
Power [kW]
2
Electric
0
0 2000 4000 6000 8000
Time, [h/year]
Figure 5.16. Duration diagram for the supply airflow rate with corresponding supply
fan electric power in Case study 1. The design airflow rate for the air
handling unit is 5,6 m3/s.
The correlation between the exhaust airflow rate and electrical energy consumption for
the exhaust air fan was not so apparent as with the supply airflow case (see figure
5.17). The maximum measured exhaust airflow rate was 5,6 m3/s, which corresponds
exactly to the design value. The maximum measured electricity use was 5,1 kW.
56
6
rate [m³/s]
Airflow rate
Exhaust m3/s
airflow
Airflow
Electric Power
Electricity kW
5
Power [kW] 3
2
Electric
0
0 2000 4000 6000 8000
Time, [h/year]
Figure 5.17. Duration diagram for the exhaust airflow rate with corresponding fan
electric power in Case study 1. The design airflow rate is 5,6 m3/s.
The measurement results reveal that the supply and return airflow rates were not
balanced well, whereas the exhaust airflow rates were somewhat higher (see fig. 5.18).
The reason for this has not been clear so far and inaccuracy in the airflow
measurements done in the main ducts on each floor is suspected.
7
Exhaust airflow rate [m³/s]
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7
Figure 5.18. Balance between supply and exhaust airflow rates in Case study 1
during measurement period of January-December 2004. The design
airflow rate is 5,6 m3/s.
Figure 5.19 presents the measured energy consumption of the supply fan in relation
with the supply airflow rate in Case study 1. In order to determine the correlation
between the electric power and supply airflow rate, a mathematical model was tried to
fit to the measured data. According to the theory, the mathematical model that would
express the relationship between the fan energy consumption and airflow rate in a
VAV system can be written as follows (see chapter 2.5.1):
y = a + b ⋅ xc (5.1)
57
Where a, b and c are constants. A fitting procedure has been conducted with the aim to
find those values of the parameters a, b and c that would describe the experimental
data the best way. The standard way of defining the best fit is to choose the parameters
so that the sum of squares of the deviations of the theoretical curve from the
experimental points for a range of independent variables is minimum. In general a
minimum value for an equation 5.2 will be looked for through performing a series of
iterations with the computer program. With the best fit the value for χ 2
is 0.
1
χ 2 (m1 , m2 ,...) = ∑ wi [ y i − f ( xi ; m1 , m 2 ,...)] (5.2)
2
n −m i
eff
Where,
χ2
the sum of squares of the deviations of the theoretical curve from the
experimental points
yi measured values of the dependent variable y
xi measured values of the independent variable x
neff total number of experimental points used in the fitting
m total number of adjustable parameters used in the fitting
wi weights of each experimental point
6
Model:
Fan power [kW]
5 y=a + b.xc
4 Model fitting
parameters:
3 W& t = 0.79 + 0.02 ⋅ V&VAV
3
χ 2 = 0.11222
2 a = 0.787 ± 0.031
b = 0.020 ± 0.001
1 c = 3.077 ± 0.044
0
0 1 2 3 4 5 6 7
Figure 5.19. Measured supply fan power at varying airflow rates in Case study 1.
The diagram also shows the model that describes the correlation
between measured fan electric power W& t and airflow rate V& . The VAV
3
design airflow rate is 5,6 m /s.
According to the calculations, the best fitting curve to the electric power and supply
airflow relationship in Case study 1 would be as follows:
W& t = 0.79 + 0.02 ⋅ V&VAV

3
(5.3)
With the ideal control case the fan power is a function of airflow rate raised into power
of 3. The model shows that with the best fit to the experimental data the value for
58
exponent is 3. However, that good correlation in reality seems doubtful. With no
constant static pressure control in the system the value for the exponent in equation 5.3
will be close to 3 and commonly in a range of 2,5 to 2,7 [33].
Nevertheless, in a VAV system with static pressure control, the value for the exponent
is in a great extent dependent on the controlled pressure level in the system. The
higher the static pressure level is held in the system compared to the total pressure, the
smaller the exponent value gets. Therefore, in the VAV system in Case study 1, where
the pressure control is applied, the correlation shown with the equation 5.3 can not be
that good in reality. Moreover, according to the model fitting parameters presented on
the figure 5.19, the χ 2
= 0.1, which means that the model has not that good fit to the
experimental data. This can also be the reason for the correlation in equation 5.3 being
douptful.
The relationship between the exhaust fan electric power and return airflow rate of the
same air-handling system is given in figure 5.20. Here the experimental data do not
show that clear relationship between the electricity use and airflow rates. Also, as can
be seen from the figure, there is a small group of electricity values before 5m3/s that
seem to be independent of airflow rates. Here the measuring errors are suspected.
6
Model:
y=a + b.xc
Fan Power [kW]
4 Model fitting
W& t = −0.48 + 0.33 ⋅ V&VAV
1.66
parameters:
3 χ 2 = 0.08912
2 a = -0.483 ±0.044
b = 0.325 ± 0.016
1 c = 1.656 ± 0.026
0
0 1 2 3 4 5 6 7
Figure 5.20. Measured exhaust fan power at varying airflow rates in Case study 1.
between measured fan electric power W& t and airflow rate V&VAV . The
design airflow rate is 5,6 m3/s.
According to the calculations, the best fitting curve to the electric power for the fan
and exhaust airflow relationship would be as follows:
W& t = −0.48 + 0.32 ⋅ V&VAV

1.66
(5.4)
Here the negative value for the first constant in equation 5.4 is not possible in the real
case. The value for this constant depends on the controlled static pressure level in the
system and will always be a positive number. The reason for the model giving the
59
negative result for this constant may be due to not so clear correlation in the
experimental data. The computer program looks just looks for the best fit possible.
Case study 2A
Figure 5.21 shows the supply airflow rates as a function of outdoor temperatures for
the building in Case study 2A. As it can be seen from the figure, the design airflow
rate 5,0 m3/s was never reached during the one-year measurement period. The
maximum of all supply air diffusers, which is 4,2 m3/s, was exceeded only one hour
during the whole monitoring period. Moreover, the supply airflow rates were in the
same range all year, regardless the outdoor temperature. Some increase in airflow rates
can be seen when the outdoor temperature raises above +15°C.
5
0
-10 -5 0 5 10 15 20 25 30

Figure 5.21. Supply airflow rate as a function of outdoor temperature in Case study
2A.The design airflow rate is 5,0 m3.
The duration diagram for the supply airflow rates and corresponding electrical energy
consumption for the supply air fan is given in figure 5.22. The total working hours of
the system was approx 7200 hours/year. The maximum measured supply airflow rate
was 4,4 m3/s and maximum electric power for supply fan was 3,0 kW.
As can be seen from the diagram, during the majority of operating time, approx 5000
hours, the system operated below 45% of the maximum airflow rate of all supply air
diffusers, which is 4,2 m3/s. As was shown previously on figure 5.21 the supply
airflow rates were distributed quite evenly over the wide range of outdoor
temperatures. This could have been affected by the use of the lecture hall and
conference rooms, which gives a quite big part to the total airflow rate in the system
(approx. 90% of the total airflow) and therefore determines the airflow use and energy
consumption to a big extent. During the summer time when the heat gains in the rooms
should be on the peak level, most of these rooms are sparsely used due to the vacation
period.
60
6 Supply
Supply airflow, m3/sm3/s
airflow rate
rate [m³/s]
Airflow
Electricity
Electric kW kW
Power,
5
3
Power [kW]
2
Electric
0
0 2000 4000 6000 8000
Time, [h/year]
Figure 5.22. Duration diagram for the supply airflow rate with corresponding fan
energy use in Case study 2A during the measurement period of May
2004-April 2005. The design airflow rate is 5,0 m3/s.
Figure 5.23 presents the duration diagram for the exhaust airflow rates and
corresponding electric power values for the exhaust air fan. The maximum measured
exhaust airflow rate was 5,3 m3/s and corresponding electricity value 5,7 kW. The
higher values of return airflow rates correspond to the occasion where the extra heat
from the condenser in the water chiller system is removed by exhaust air and therefore
the return air fan is running on the maximum speed.
The balance diagram between the supply and exhaust air for the same building shows
that the supply and exhaust airflow rates were well balanced (see figure 5.24). The
points above the diagonal line correspond again to the situations, where the water
chiller system was dumping the heat from the condenser.
6 3
rate [m³/s]
Return
Exhaustairflow ratemm3/s
airflow, /s
Airflow
Electricity kW kW
Electric Power,
5
3
Power [kW]
2
Electric
0
0 2000 4000 6000 8000
Time, [h/year]
energy consumption in Case study 2A during the measurement period
of May 2004-April 2005. The design airflow rate is 5,0 m3/s.
61
5

4
0
0 1 2 3 4 5
Figure 5.24. Balance between supply and exhaust airflow rates in Case study 2A
during the measurement period from May 2004 - April 2005. The
design airflow rate is 5,0 m3/s. The points above the diagonal line
correspond to the situation, where the water chiller system was dumping
the heat from the condenser.
Figure 5.25 presents the energy consumption of the supply air fan at varying airflow
rates. According to the calculations, the best fitting curve to the electric power of the
supply fan and supply airflow rate correlation would be as follows:
W& t = 0.18 + 0.21 ⋅ V&VAV

1.73
(5.5)
5
Model:
Fan power [kW]
y=a + b.xc
4
Model fitting
3 parameters:
W& t = 0.18 + 0.21 ⋅ V&VAV
1.73
χ 2 = 0.00237
2
a = 0.180 ±0.005
1
b = 0.211 ± 0.003
c = 1.729 ± 0.010
0
0 1 2 3 4 5

Figure 5.25. Measured supply fan power at varying airflow rates in Case study 2A.
design airflow rate is 5,0 m3/s
Figure 5.26 shows the measured exhaust fan power at varying airflow rates. The best
fitting curve to the electric power of the exhaust fan and exhaust airflow rate
correlation would be as follows:
62
W& t = 0.36 + 0.26 ⋅ V&VAV
1.86
(5.6)
5
Model:
Fan Power [kW]
4 y=a + b.xc
W& t = 0.36 + 0.26 ⋅ V&VAV
1.86
3 Model fitting
parameters:
2 χ 2 = 0.0039
a = 0.362 ± 0.006
1 b = 0.259 ± 0.003
c = 1.863 ± 0.007
0
0 1 2 3 4 5

Figure 5.26. Measured exhaust fan power at varying airflow rates in Case study 2A.
Case study 2B
Figure 5.27 shows the measured airflow rates presented as a function of the outdoor
temperatures in Case study 2B. When in the A building in Case study 2 the supply
airflow rates were somewhat increasing with the outdoor temperatures rising above
+15°C (see figure 5.21), then in this building part similar tendency did not occur. The
airflow rates were in the same range of 1,0 – 2,0 m3/s all year around.
The duration diagram of the supply airflow rates and corresponding electricity use of
the supply air fan in the same building is presented in figure 5.28. The total operating
hours of the system was approx 4900 hours/year.
The design airflow rate is 3,6 m3/s, meanwhile the maximum sum airflow of supply air
diffusers is 3,0 m3/s. As shown in the figure, the system never reached the design
airflow rate 3,6 m3/s during the measurement period. Moreover, it operated with less
than 45% of the maximum airflow rate of all supply air diffusers during 80% of the
operating hours, about 3900 hours/year. The maximum measured supply airflow rate
was 2,3 m3/s, which corresponds to approx 64% of the design airflow rate and the
maximum measured supply fan energy effect was 1,4 kW.
These low values compared to the design values can be contributed to a low use of the
rooms. According to research carried out in office buildings, the actual use of office
rooms has stated to be about 50% during working hours [36].
63
4

3
0
-10 -5 0 5 10 15 20 25 30
Figure 5.27. Supply airflow rate as a function of outside temperature in Case study
2B. The design airflow rate is 3,6 m3/s.
4 Supply
Supply airflow
airflow,rate
m3/sm3/s
rate [m³/s]
Airflow
Electric power,
Electricity kW kW
3
2
Power [kW]
Electric
0
0 2000 4000 6000 8000
Time , [h/ye ar]
Figure 5.28. Duration diagram for the supply airflow rate with corresponding fan
energy use in Case study 2B during the measurement period of May
2004 – April 2005. The design airflow rate is 3,6 m3/s.
The measured exhaust airflow rates were in the same range as the supply airflow rates,
which imply good balance conditions between the supply and exhaust air. The
maximum measured exhaust airflow rate was 3,3 m3/s and maximum energy effect for
exhaust air fan was 3,1 kW (see fig. 5.29).
Figure 5.30 shows the balance between the supply and exhaust airflow rates.
Somewhat higher maximum values compared to the supply airflow are due to the
water chiller condenser heat, which is dumped to the exhaust air, as it is also the case
in the air-handling system in Case study 2A.
64
4 Return
Exhaust airflow 3
ratemm3/s
airflow, /s
rate [m³/s]
Airflow
Electric power,
Electricity kW kW
2
Power [kW]
Electric
0
0 2000 4000 6000 8000
Time, [h/year]
energy use in Case study 2B during the measurement period of May
2004 – April 2005. The design airflow rate is 3,6 m3/s.
4
Exhaust airlfow rate [m³/s]
0
0 1 2 3 4
Figure 5.30. Balance between the supply and exhaust airflow rates in Case study 2B
during the measurement period from May 2004 - April 2005. The
Figure 5.31 presents the energy consumption of the supply air fan at varying airflow
rates. According to the calculations, the best fitting curve to the electric power of the
supply fan and supply airflow rate correlation would be as follows:
W& t = −0.12 + 0.45 ⋅ V&VAV

1.34
(5.7)
According to the model fitting parameters presented in2the figure 5.31, the model
seems to fit to the data rather well. The value for the χ is relatively low, meaning a
good fit of the model to the experimental data. However, here the negative value for
65
the constant in equation 5.7 is not possible in the real case. The value is dependent on
the controlled static pressure level in the system and will always be a positive number.
Therefore the correlation expressed with the equation 5.7 can be questioned, especially
at low airflow rates.
Model:
Fan power [kW]
2 y=a + b.xc
W& t = −0.12 + 0.45 ⋅ V&VAV
1.34
Model fitting
parameters:
1
χ 2 = 0.00065
a = -0.122 ±0.016
b = 0.450 ± 0.014
c = 1.340 ± 0.028
0
0 1 2 3 4

Figure 5.31. Measured supply fan power at varying airflow rates in Case study 2B.
Measured exhaust fan power at varying airflow rates can be seen in figure 5.32. The
best fitting curve to the electric power of the exhaust fan and exhaust airflow rate
correlation would be as follows:
W& t = 0.14 + 0.30 ⋅ V&VAV

1.92
(5.8)
Model:
Fan power [kW]
2 W& t = 0.14 + 0.30 ⋅ V&VAV

1.92 y=a + b.xc
Model fitting
parameters:
χ 2 = 0.00141
1
a = 0.135 ± 0.011
b = 0.297 ± 0.007
c = 1.916 ± 0.018
0
0 1 2 3 4

Figure 5.32. Measured exhaust fan power at varying airflow rates in Case study 2B.
between the measured fan electric power W& t and airflow rate V&VAV . The
66
In general, the correlations between airflow rate and fan power in different air
handling systems in both case studies revealed that the fan power is a function of
airflow raised to a power, which had a value less than 2. In a VAV system with
adjustable speed drive control with frequency converter, this exponent is mainly
dependent on the proportion of duct static pressure set-point level compared to the
total fan pressure. The higher the static pressure control level is compared to total
pressure, the smaller the value for the exponent gets and the more linear is the
relationship between the fan power and airflow rate [33, 49].
However, it should also be noted that in reality the system will not follow the
theoretical curves, expressed with the correlations, over the whole airflow range. It is
most likely that the parameters of these correlations will change when the airflow rate
decreases. This is due to the change in both system and fan characteristics.
The power demand of the fan depends besides the airflow rate also on the total
pressure rise and the total efficiency of the fan (see equation 2.3, chapter 2.5.1). Both
of these parameters change when the airflow is decreased. The total fan efficiency of
the fan goes down when the airflow gets below approx 40% of the design value, as
was shown in figures 2.9 and 2.10. Furthermore decreasing the airflow can lead to a
change in airflow conditions in the system from turbulent to laminar, which will result
to a change in system characteristics. Therefore the exact fan power-airflow
correlation is hard to evaluate in practice when the airflow decreases considerably
from the design airflow rate. This explains also the results from fitting the model to the
experimental data in both case studies. For example in Case study 1, the correlation for
the supply fan power and airflow rate did not show the best fit and the value for the
exponent in equation 5.3 seemed doubtful. Also the negative values for the constants
in equations 5.4 and 5.7 can be related to the problem described above.
5.1.6 Specific Fan Power- SFP
Electrical efficiency of a ventilation system is commonly measured by Specific Fan

Power (SFP) value. It is calculated by dividing the electric power demand of all the
fans in the air distribution system by the highest value of supply air flow rate or
exhaust air flow rate through the building under design load conditions [49] (see
equation 2.8, chapter 2.5.1)
It is recommended that the SPF value for a new building project should not exceed 2,0
kW/(m3/s) and for rebuilding and conversion work the maximum SFP 2,5 kW/(m3/s)
should be aimed [7, 47].
Figure 5.33 presents the calculated SFPVAV values given as a function of measured
maximum hourly airflow rates in Case study 1 during the measurement period of
January to December 2005. According to the figure, the SFP value at design airflow
rate 5,6 m3/s is approx 1,6 kW/(m3/s). This SFP value is quite good for an existing
building and is probably due to the preserved duct system and that the air handling
unit was originally selected for higher airflow rate.
67
2,0
Specific Fan Power SFP

Model:
1,5 y=a + b.xc
[kW/m³/s] SFPVAV = 0.41 + 0.07 ⋅ V&VAV

1.65
Model fitting
1,0 parameters:
χ 2 = 0.03195
0,5 a = 0.412 ± 0.020

b = 0.067 ± 0.007
c = 1.650 ± 0.052
0,0
0 1 2 3 4 5 6 7
Maximum hourly airflow rate [m³/s]
Figure 5.33. Specific Fan Power given as a function of maximum hourly airflow rate
in Case study 1. The diagram also shows the model that describes the
correlation between the calculated SFPVAV value and measured
maximum hourly airflow rate V&VAV . The SFP value at the design airflow
rate 5,6 m3/s is approx 1,6 kW/(m3/s).
Even though the airflow rates were not well balanced in the VAV system in Case
study 1 and according to figure 5.33 the relationship between the SFPVAV value and
airflow rate is not so clear, the specific fan power tends to decrease with decreased
airflow rates. According to the calculations, the best fitting curve to the SFPVAV value
and maximum hourly airflow rate correlation would be as follows:
SFPVAV = 0.41 + 0.07 ⋅ V&VAV

1.65
(5.9)
Figure 5.34 and 5.35 present the calculated SFPVAV values for the buildings in Case
study 2. The airflow rate values correspond to the measured values during the
measurement period of May 2004 to April 2005.
The results from the calculations for the air handling system in Case study 2A, shown
in figure 5.34, yield that the majority of the SFPVAV values are between 1,0-1,5
kW/(m3/s). When calculating the SFP value for the design airflow conditions with 5,0
m3/s, the results yield the SFP value to be approx. 1,7 kW/(m3/s). This result is below
the designed value 2,1 kW/(m3/s), presented in table 3.2 in chapter 3.2.
According to the calculations, the best fitting curve to the SFPVAV value and maximum
hourly airflow rate correlation for the air-handling system in Case study 2A would be
as follows:
SFPVAV = 0.54 + 0.27 ⋅ V&VAV

0.89
(5.10)
68
Specific fan power SFP [kW/m³/s]
2,0
SFPVAV = 0.54 + 0.27 ⋅ V&VAV

0.89
1,5 Model:
y=a + b.xc
1,0 Model fitting

parameters:
χ 2 = 0.00368
0,5 a = 0.539 ± 0.025
b = 0.267 ± 0.020
c = 0.890 ± 0.036
0,0
0 1 2 3 4 5
in Case study 2A. The diagram also shows the model that describes the
maximum hourly airflow rate V&VAV . The SFP value at the design airflow
rate 5,0 m3/s is approx 1,7 kW/(m3/s).
According to figure 5.35, in Case study 2B the SFPVAV values seem to vary also
linearly. However, according to the calculations, the best fitting curve to the SFPVAV
value and maximum hourly airflow rate correlation would show different results:
SFPVAV = −9.96+10.51⋅V&VAV
0.05
(5.11)
Specific fan power SFP [kW/(m³/s)]
2,0
1,5 SFPVAV = −9.96 + 10.51⋅ V&VAV

0.05
Model:
y=a + b.xc
Model fitting
1,0 parameters:
χ 2 = 0.00229
0,5 a = -9.959 ±17.815
b = 10.510±17.805
c = 0.054 ± 0.088
0,0
0 1 2 3 4
in Case study 2B. The diagram also shows the model that describes the
maximum hourly airflow rate V&VAV . According to the model the SFP
value at the design airflow rate 3,6 m3/s would be approx 1,3
kW/(m3/s).
69
The negative value for the first constant in equation 5.11 is not possible in reality.
Moreover, as can be seen from the model fitting parameters in figure 5.35, the fitting
results give a very big inaccuracy for the constants a, b and c. Therefore the fit of the
model in this case is doubtful.
However, according to the model the SFP value becomes 1,3 kW/(m3/s) with the
dimensioned airflow rate 3,6 m3/s (the designed value for SFP is 1,9 kW/(m3/s)). This
value is also very low and can be questionable since the model did not show the
reasonable fit results.
In general, despite the questionable results from the model fit for Case study 2B, the
SFPVAV values are relatively low, showing that the VAV systems worked energy
efficiently.
Figure 5.36 gives the SFP values at average airflow rates (SFPA) of VAV systems in
different Swedish office buildings together with the data measured from case studies 1
and 2. The data is from the energy auditing funded by the Swedish Energy Agency in
2005 [17]. All together the energy use of 123 office buildings was analysed. The figure
5.36 presents the SFPA values for 43 air-handling systems, which work with variable
air volume flow. The figure also gives the data about the construction year of the
building. However, the renovation year of the air handling systems is not known.
As can be seen from the figure, the SFPA values for the average airflow rates in Case
study 1 and 2 are considerably lower than the SFPA values for other office buildings
with VAV systems. In Case study 1 the average airflow rate per year was 4,6 m3/s and
the corresponding SFPA value is 1,3 kW/(m3/s). For the Case study 2A these numbers
were 2,2 m3/s and 1,1 kW/(m3/s) and for the Case study 2B 1,8 m3/s and 0,9
kW/(m3/s).
70
Construction year
of the building
10,0 1890
1920
Specific Fan Power, SFP
8,0 1929
1940
1955
[kW/(m³/s)]
6,0 1960
1969
4,0 Case study 1 1972
1973
2,0 1976
1988
Case study 2 1999
0,0
2002
0 2 4 6 8 10 12 14 16 2004
Av e rage airflow rate , [m³/s]
Figure 5.36. Specific Fan Power at average airflow rates of VAV systems in
different Swedish office buildings. The data is from the energy auditing
funded by the Swedish Energy Agency in 2005 [17]. The figure shows
also the construction year of the building itself. The renovation year of
older buildings is not known. The SFPA values for the average airflow
rate of 4,6 m3/s in Case study 1 was 1,3 kW/(m3/s), for Case study 2A
these numbers are 2,2 m3/s and 1,1 kW/(m3/s); for Case study 2B 1,8
m3/s and 0,9 kW/(m3/s).
71
5.2 Results from the laboratory measurements
The measured values of temperatures and air velocities in different room points are
presented in tables A.1 and A.2 in APPENDIX 2. The values in the tables are average
values of three replicates made in the same conditions. The tables also give the values
for calculated draught rates in these points.
According to the measured data, the room air temperature was relatively uniform over
the measured levels for all airflow rates. However, there seemed to be some
differences (up to 0,7°C) between the temperatures on the level of 0,1m compared to
the level of 1,1m, where the temperatures were somewhat higher. Here the somewhat
cold test room floor may have an effect on the results.
The mean air speed and draught rating distributions at different supply airflow rates in
two mounting cases, with and without suspended ceiling, are presented in figures 5.37
and 5.38. The figures show the percentage of the measured points being in the
specified range of air velocity and draught rate values.
According to thermal comfort guidelines, the air velocity in the occupied space should
not exceed 0,15 m/s [8, 52] and the draught rating is limited to 15% [30]. From the
measurement results it occurs that these limits were exceeded in few measured points
in the room and that mainly at maximum airflow rates. From the figure 5.37 can be
seen that the average air velocity in the majority of the measured points was in a range
of 0,06-0,15 m/s with the maximum airflow rate 50 l/s. With the medium supply
airflow rate 25 l/s this range was 0,06-0,10 m/s and when the airflow was set to
minimum 10 l/s the air velocities were less than 0,05 m/s.
There seem to be no substantial difference in the results with and without suspended
ceiling mounting case. However, with the suspended ceiling, the number of measured
points, where the air velocities and draught rating exceeded the comfort conditions,
was somewhat higher compared with no suspended ceiling case.
72
Without suspended ceiling With suspended ceiling
Without false ceiling With false ceiling
10 l/s 10 l/s
measured points, % 100
100 100
100
Fraction of total
80
80 80
80
60
60 60
40
40 40
20
20 20
20
00 0
0
<0,05
<0,05 0,06-0,10
0,06 - 0,11-0,15
0,11 - >0,15
>0,15 <0,05
<0,05 0,06-0,10
0,06 - 0,10 0,11-0,15
0,11 - 0,15 >0,15
>0,15
0,10velocity,
Air 0,15m/s
Air velocity, m/s
25 l/s 25 l/s
measured points, %
100
100 100
100
Fraction of total
80
80 80
80
60
60 60
60
40
40 40
40
20
20 20
20
00 00
<0,05
<0,05 0,06 - 0,10
0,06-0,10 0,11 - 0,15
0,11-0,15 >0,15
>0,15 <0,05
<0,05 0,06 - 0,10 0,11
0,06-0,10 - 0,15
0,11-0,15 >0,15
>0,15
Air velocity, m/s Air velocity, m/s

50 l/s 50 l/s
measured points, %
100
100 100
100
Fraction of total
80
80 80
80
60
60 60
60
40
40 40
40
20
20 20
20
00 00
<0,05
<0,05 0,06-0,10
0,06 0,11-0,15
- 0,10 0,11 - 0,15 >0,15
>0,15 <0,05
<0,05 0,06-0,10
0,06 - 0,10 0,11-0,15
0,11 - 0,15 >0,15
>0,15
Air velocity, m/s Air velocity, m/s
Figure 5.37 Air speed distributions in the test room with different supply airflow
rates and with false ceiling mounting case. The supply air temperature
was +15°C. According to thermal comfort guidelines [8, 52] the air
velocity in the occupied space should not exceed 0,15 m/s.
73
Without suspended ceiling With
With suspended ceiling
suspended ceiling
Without suspended ceiling
10 l/s 10
10l/sl/s
10 l/s
%
total 100
100 100
100
m easured points,
points,
points,
Fraction of total
Fraction of total
80
80 8080
easured of
60
60 6060
[% ]
[% ]
mFraction
measured
40
40 4040
20
20 2020
0 00
0
<5 5 - 10 10 - 15 >15 <5 5 - 10 10 - 15 >15
Draught rating
Draught ratingDR,
DR,[%]
% Draughtrating
Draught ratingDR,
DR,[%]
%
25 l/s 25 l/s
25 l/s 25 l/s
%
100 100
m easuredofpoints
100
m easured points, [% ]
100
Fraction of total
total
points,
Fraction of total
80
80 80
measured[% ]
60
60 60
Fraction
40
40 40
20
20 20
20
00 0
0
<5 5 - 10 10 - 15 >15 <5 5 - 10 10 - 15 >15
Draughtrating
Draught ratingDR,
DR,%
[%] Draught
Draughtrating DR,DR,
rating [%] %
50 l/s
50 l/s 50 l/s
50 l/s
%
100
100 100
100
of total
points,
m easured points,
Fractionpoints,
Fraction of total
Fraction of total
80
80 80
80
60
60 60
60
[% ]
measured
[% ]
m easured
40
40 40
40
20
20 20
20
00 00
<5 5 - 10 10 - 15 >15 <5 5 - 10 10 - 15 >15
Draught rating DR, [%] Draught rating
Draught rating DR, % Draught ratingDR,
DR,[%]%
Figure 5.38 Draught rating distributions in the test room with different supply
airflow rates and mounting cases. The supply air temperature was
+15°C. According to thermal comfort guideline ISO7730 the draught
rating is limited to 15% [30].
74
For analysing more in detail if the measured values of air velocities in the occupied
space depend on different parameters that were varied during the experiment, an
analysis of variance (ANOVA) was carried out with the measurement results. The
procedure of ANOVA employs the statistic F to test the statistical significance of the
differences among the obtained mean values of two or more random samples from a
given population. Here the effect of different parameters such are the room point, the
level of a measured point, the ceiling and the airflow rate on the average air velocity in
the room was studied. The results of the ANOVA test revealed a statistical
significance of an effect of each parameter and whether the variability among the
measured mean air velocities is due to experimental error or not.
In order to statistically identify parameters that have an influence on the mean air
velocity in the occupied space, a simple mathematical model was built up and tested
with ANOVA. The simplest such model presupposes that an observation y can be
represented as a sum of a mean value, all the possible effects and an error as follows:
y ijklm = y + β i + α j + τ k + γ l + ε ijklm (5.12)
Where,
yijklm a single observation of an air speed. The index denotes the combination of
different variables listed below. The index m marks the replicate number and
varies from m=1…3 (every airflow case was done in 3 replicates)
y general mean of an air speed
βi a room point effect. Index i varies from i = 1…9 (the measurements were done
in nine room points)
αj a room level effect. Index j varies from j = 1…3 (the measurements were done
on three different room levels: 0,1m, 0,6m, 1,1m)
τk a ceiling effect. Index k varies from j = 1…2 (the measurements were done
with two different mounting cases: with and without false ceiling)
γl an airflow effect. Index j varies from j = 1…3 (the measurements were done
with three different airflow cases: 50l/s, 25l/s, 10l/s)
εijkl an error
With this model the main effects of different variables can be tested. This can identify
the influence of separate variables on the mean air velocity in the room. The
preliminary assumption for the validity of this model is that the errors εijkl are
independently and identically distributed in a normal distribution with mean zero and
variance σ2.
Each of the variables in the model was tested for the zero hypothesis H0, which states
that there is no effect from that variable, e.g. H0: βi = 0. If there were an effect of a
variable to the mean air velocity the hypothesis would state the opposite, e.g. H1: βi ≠
0.
The chosen confidence level in the analysis accounted here is 95%, meaning that the
conclusions have a 5% chance of not being true (p = 0,05). For example, if the
probability that the calculated F value exceeds the tabulated value given for the F-
distribution in the F-table is equal or less than p = 0.05, then there is at least 95%
chance that there is an effect of a single variable to the mean air velocity. Thus the
zero hypothesis is disregarded.
75
It was preliminary assumed that the airflow has an effect to the mean air velocity in
the occupied zone and the results from the analysis of variance revealed the same. The
probability that the variability of the mean air velocity values with different airflow
rates can be attributed to experimental error is P(F>577,85) = 2,2.10-16. This low value
of probability means that the airflow rate has a statistically significant effect on mean
air velocities in the occupied space.
The main effects also occur from the room level, P(F>15,29) = 0,0001, and the ceiling
P(F>15,29) = 0,0026. This means that the room level and whether the supply air
device is mounted below the ceiling or above it, as in the suspended ceiling mounting
case, affect the mean air velocities in the occupied space.
For analysing more in detail the effect of different variables, a similar model to
equation 5.12 was tested for each airflow case separately. The results are presented in
table 5.5.
Table 5.5
Statistically significant effects of different variables on the mean air velocity in the
occupied space.
Main effect Combined all 10 l/s 25 l/s 50 l/s
airflows
Room point NO NO YES YES
Room level YES YES NO YES
Ceiling YES YES YES NO
Airflow rate YES - - -
As can be seen from the table 5.5 the level and ceiling had single effects when the
airflow was 10 l/s. For the level the probability is P(F>60,47) = 9,005.10-13 and for the
ceiling P(F>63,05) = 3,52.10-13.
The results from the second airflow case 25 l/s did not indicate any single effect of a
room level. However, here the ceiling had an effect P(F>17,65) = 4,42.10-5 and also
the room point P(F>5,05) = 0,025.
The results from maximum airflow case 50 l/s revealed single effects from both the
room point P(F>4,79) = 0,030 and room level Pr(F>6,45) = 0,012.
The interaction effects of different parameters were also tested, but no interaction
effects were revealed from the results.
The results of ANOVA tests made with the model that was described above do not
show the size of a single effect, e.g. if the room level has a higher effect compared to
ceiling mounting case. Moreover, the direction of the variation is not known, e.g. in
which case the highest results appear. They show statistically if different parameters
affect the results or if the variation is mainly due to experimental error.
In order to further analyse the effect of different parameters on mean air velocity in the
room, iso-velocity profiles were composed for different measurement cases. The
results are presented in APPENDIX 2. As can be seen from the iso-velocity figures,
the room points where the required draught rating and air velocity values were
exceeded were all locating on one side of the room (measuring points 3, 6, 9). This
76
was the empty side of the test room. Moreover, the most critical point, room point nr
6, was locating on the level of 0,1m above the floor. The upper levels of the same
measuring point do not have any higher velocities.
Figure 5.39 presents the standard deviation as a function of the mean air velocity with
all different airflow conditions at three measurement levels for no suspended ceiling
installing case. It can be seen that the fluctuation of the air velocity was increasing
when the average air velocity in measured points increased. In addition, the gradient
of the regression lines decreased slightly as the measuring level decreased from 1,1m
to 0,1m. This implies that higher fluctuation of the air velocities was found at ankle
level compared to other measured levels. Similar data has also found in previous
studies about airflow characteristics [12].
0,1m
0,6m
0,08
Standard deviation, m/s
y = 0,3824x 1,1m
0,07 R2 = 0,8265
Linear (0,1m)
0,06
Linear (0,6m)
0,05 Linear (1,1m)
0,04 y = 0,3285x
R2 = 0,7915
0,03
0,02 y = 0,3335x
R2 = 0,8019
0,01
0
0 0,05 0,1 0,15 0,2 0,25
mean air speed, m/s
Figure 5.39 Standard deviation of air velocity with three different airflow rates and
at different levels in the test room. The diagram corresponds to no
suspended ceiling mounting case. The supply air temperature was
+15°C.
Since the processes in the room regarding airflow dynamics are complex and
influenced by many different factors, it is hard to derive definite conclusions why the
risk of draught is occurring in the described points in the room. According to the
measurements, more draught risk was indicated with the false ceiling case, even
though the false ceiling set-up increased the height under the supply air device
compared to without false ceiling mounting.
In addition, all the heat sources in the room have an influence to the air motion in the
room by giving rise to a buoyancy induced velocity scale which can match the
velocities generated by a jet in the occupied zone [18]. This could be an explanation
why all the critical points occurred on the empty side of the test room, since the only
influencing heat source was the ceiling luminance. There was no draught risk
indicated in the working place in the test room (left side of the room).
In general more research should be conducted on this issue in order to make definite
conclusions.
77
78
6 EVALUATION OF THE HEAT GAINS IN A
VAV DUCT SYSTEM
6.1 Mathematical calculation of the temperature
rise in the duct system
The cold air streams in ducts are influenced by heat gains due to the temperature
difference between the supply air in the duct and the room temperature. The amount of
heat transferred through the duct depends on the temperature difference, the properties
of the insulation material and its thickness. It also depends on airflow rates in the duct
and varies under different airflow conditions. In order to evaluate the supply air
temperature and cooling capacity of supply air under different operating conditions, it
is important to estimate the total heat gains to the system.
Figure 6.1 illustrates an air duct. The change of temperature of a ducted air stream
under the influence of a heat gain can be evaluated from a heat balance equation as
follows:
− dQ& c + q& tr ⋅ dA = 0 (6.1)
Where,
dQ& c change in the heat capacity of the air flowing along the duct, [W]
q& tr transmission heat loss, [W/m2]
dA surface area of a duct element, [m2]
U t0 dA t,°C t0
L0 M& aa dx
∆t2 tx= t(x) ∆t1
tin tx= t(x) tin

x
duct length, m x
0 L 0 L
Figure 6.1. An example of an air duct and the temperature change inside the duct.
From the figure: L0 –perimeter of the duct, U- Thermal transmittance
between the air in the duct and outside temperature, tin(x) – supply air
temperature, t0 – Outside temperature, t(x) - steady state air temperature
along the duct, M& a – air mass flow rate in the duct, L- duct length.
The heat capacity flow of the air in the duct can be written:
dQ& c = C& ⋅ dt x (6.2)
Where,
C& heat capacity of air, [W/°C]
dtx steady state air temperature along the duct, [°C]
79
The transmission heat loss though the duct element dx (see fig. 6.1) can be written
according to equation 6.3. It is assumed that the supply air temperature tin is lower than
outside temperature t0: tin < t0
q& tr ⋅ dA = U ⋅ (t 0 − t x ) ⋅L 0 ⋅dx [W] (6.3)
Where,
L0 perimeter area of the duct, [m]
U Thermal transmittance between the air in the duct and outside temperature,
[W/(m2 .K)]
t0 Outside temperature, [°C]
tx steady state air temperature along the duct, [°C]
For determining the tx in the equation 6.3 following assumptions are made:
• the outside temperature is constant t0 = const
• Internal temperature difference perpendicular to the airflow direction within
the channel can be neglected.
• the surface temperature of the duct is assumed to be uniform
• thermal transmittance between the air and outside temperature is U
(W/(m2°C)) can be considered constant with mass air flow rate M& a being
constant
The temperature deviation tx-t0 declines exponentially from tin- t0 to zero. The outflow
temperature and temperature change in the duct can be written as follows [26]:
L
−
t x = t 0 + (t in − t 0 ) ⋅ e lc
(6.4)
L L
− −
∆t1 = t 0 − t in + (t in − t 0 ) ⋅ e Lc
= ∆t 2 (1 − e Lc
) (6.5)
Where,
∆t1 temperature change in the duct, ∆t1= tx-tin, [°C]
∆t2 temperature difference between the inlet air temperature and outside
temperature,
∆t2= t0- tin ; [°C]
tin the inlet air temperature at x=0, [°C]
t0 the outside temperature, [°C]
tx the supply air temperature at x = L, [°C]
Lc characteristic length, [m]
L duct length, [m]
The length Lc (m) in eq. 6.4 is a characteristic length for the interaction between
convective heat flow along the duct and the transverse heat loss and is defined:
M& a ⋅ c pa
Lc = [m] (6.6)
U ⋅ L0
80
Where,
M& a air mass flow rate in the duct, [kg/s]
cpa specific heat capacity of air of constant pressure, [J/(kg.K)]
The value of characteristic length lc is indicative of the rate of response of a single

capacity system to a sudden change in the environmental temperature. When the duct
length is equal to the ratio given with the equation 6.6, the temperature difference tx-tin
is equal to 63,2% of the initial difference t0-tin (see figure 6.2).
t,°C t0
tx
t0 -tin
63,2% tx -tin
tin
0 Distance from the

Lc entrance, m
Figure 6.2. The value of the characteristic length. The characteristic length Lc
expresses the distance when the temperature difference tx-tin is equal to
63,2% of the initial difference t0-tin.
The denominator U·L0 in the equation 6.6 is called a linear heat transfer coefficient [1]
and can be written as:
1
U ⋅ L0 = u l = (6.7)
1 l 1
+∑ +
Ai ⋅ α i λ d ⋅ Am Ao ⋅ α o
Where,
U total thermal transmittance of a duct, [W/(m2 .K)]
ul linear thermal transmittance of a duct, [W/(m .K)]
αi convective heat transfer coefficient between the air and inner duct surface,
[W/(m2°C)]
αo heat transfer coefficient between the outside air and outside duct surface,
[W/(m2 .K)]
λd thermal conductivity of the duct layer, [W/(m .K)]
l thickness of a duct layer, [m]
ai inner perimeter area of a duct, [m]
ao outer perimeter are of a duct, [m]
am logarithmic middle area of a duct layer, [m]
a 2 − a1
am = (6.8)
a
ln 2
a1
Where a1 and a2 are inner and outer areas of a duct layer, [m]
81
For circular ducts, equation 6.7 can be written:
1
U ⋅ L0 = u l = n −1
(6.9)
1 1 r 1
+∑ ln k +1 +
2πriα i k =1 2πλdk rk 2πroα o
Where,
ri inner radius of the duct, [m]
ro outer radius of the duct, [m]
The linear thermal transmittance ul is dependent on the flow conditions inside the duct
and remains relatively constant with the air velocities and duct sizes in common range.
However, a decrease of air velocity can lead to a change in the flow conditions in a
way that the turbulent flow inside the duct is changed to laminar flow. This leads to a
considerable change in ul value and therefore should be further analysed for evaluating
the duct heat gains.
The total thermal transmittance U:
1 [W/(m2 .K)] (6.10)

U=
1 l 1
+∑ +
αi λd αo
Where,
l thickness of the duct layer, [m]
In practical calculations the thermal conductivity of a duct layer λd is considered only

for the duct insulation, where the λi ≈ 0,04 W/(m.K). The resistance of the metal is
ignored, since the thermal conductivity is approx λm ≈ 200-500 W/(m.K) and gives a
very little impact to overall thermal transmittance.
The heat transfer coefficient between the outside air and outside duct surface αo is
dependent on the location of the duct and depending on both the convective heat
transfer and heat transfer by radiation between surrounding surfaces. Most commonly
the ducts are placed in the shafts and above false ceilings and in these cases the
convective heat transfer between the outside air and outside duct surface is by natural
convection.
However, in practice the values for αo are difficult to establish with any certainty,
since the proximity of the duct to a ceiling or wall has an inhibiting effect on heat
transfer and tends to increase the value of αo. A value of αo=10 W/(m2 .K) is suggested
[38]
and this value considers both the convective and radiation part of heat transfer.
For determining the convective heat transfer coefficient between the air and inner duct
surface αi, equation 6.11 can be used:
Nu ⋅ λ air
αi = [W/(m2 .K)] (6.11)
Dh
82
Where,
αi convective heat transfer coefficient between the air and inner duct surface,
[W/(m2 .K)]
Nu Nusselts number
Dh hydraulic diameter of the duct, [m]. For circular ducts it is equal to the inside
diameter of the duct.
λ air thermal conductivity of air, [W/(m.K)] For the dry air with temperature +20°C
the thermal conductivity of air is λ air= 26,03.10-3 W/(m.K).
The Nusselt number depends primarily on the flow condition, which is characterized
by the Reynolds number Re and can be calculated according to equations 6.12, 6.14
and 6.16 [1, 21].
For Re>2300 (from transitional region to fully turbulent flow region):
0.125 ⋅ f (Re− 1000) Pr

Nu tc = (6.12)
1 + 12.7(0.125 ⋅ f ) 0.5 ⋅ (Pr 2 / 3 − 1)
Where,
f = (0.79 ln Re − 1.64) −2 (6.13)
For Re<2300 (laminar flow):
⎡ ⎛ Re⋅ Pr⋅ Dh ⎞ ⎤
⎢ 0.067 ⋅ ⎜ ⎟ ⎥
⎢ ⎝ L ⎠ ⎥ (6.14)
Nu lc = 3.65 +
⎢ ⎛ Re⋅ Pr⋅ Dh ⎞ ⎥⎥
2/3
⎢ 1 + 0.045 ⋅ ⎜ ⎟
⎣⎢ ⎝ L ⎠ ⎦⎥
Where,
Nulc Nusselt number for laminar forced convection
Nutc Nusselt number for turbulent forced convection
Pr Prandtl number. For the dry air with temperature +20°C the Prandtl number
is Pr=0,7
L length of the duct
n = 0,4 if the duct surface temperature > air temperature in the duct (heating the air)
n = 0,3 if the duct surface temperature < air temperature in the duct (cooling the air)
Re Reynolds number.
The Reynolds number:
v ⋅ Dh
Re = (6.15)
υ
Where,
v air velocity of the airflow in the duct, m/s
υ kinematic viscosity of air, m2/s. For the dry air with temperature +20°C the
kinematic viscosity is υ =15,13.10-6 m2/s
83
Since in the transitional region the Nusselt number increases rapidly and it can be hard
to determine these values in the region close to the laminar flow conditions with a
good precision. Moreover, the equation 6.14 also considers the entrance effect, which
is important especially in short ducts and in laminar flow conditions. However, the
equation 6.15 does not consider this effect when the Reynold number values are close
to 2300. In order to establish more stable relationships for the Nusselt number in this
region, an interpolation can be done for the conditions where 2000 < Re > 3000. The
correlation can be written as follows:
Nu trc = (1 − k ) Nu lc + k ⋅ Nu tc (6.16)
Re − 2000
k= (6.17)
3000 − 2000
Where,
Nu trc Nusselt number for the transitional region with forced convection.
As mentioned before, the ul -value is almost independent of the air velocity for the
range of duct sizes in common use. For example in a D160 diameter duct with the air
speed 1,5 m/s the ul -value with 30mm duct lagging and with λi = 0,035 W/(m.K) is
0,90 W/(m.K). For a D250 duct with the same lagging and with the air velocity 2,5 m/s
the U-value is 0,96 W/(m2°C).
However, with air velocities lower than 1,5 m/s the total ul-value decreases
considerably as can be seen from the figure 6.3. Here the change in flow conditions
has an important influence. For example in a connection duct with the diameter D250
and maximum airflow rate 50 l/s the air velocity in the duct is 1 m/s and the Reynolds
number Re≈ 16800, which corresponds to fully turbulent airflow conditions. After
decreasing the airflow in the same duct to a minimum airflow rate for example to 7 l/s,
the corresponding air velocity in the duct decreases to 0,14 m/s and the Reynolds
number to Re ≈ 3300. Here the flow conditions are in the transitional region.
ul, W/m2/K
velocities in common range

1,1
D100
1,0
Thermal transmittance U,
D160
0,9
D250
0,8
D400
0,7
0,6 D1000
0,5
0,4
0,3
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0
Air velocity, m/s
Figure 6.3. Linear thermal transmittance ul with different air velocities and duct
sizes in common use. The diagram is calculated for air ducts with
constant insulation thickness li = 30mm; λi = 0,035 W/(m2°C).
84
Figure 6.4 gives an example of calculated duct heat gain expressed as a temperature
change ∆t1 per temperature difference between the inlet and outside air: ∆t2. The
figure is calculated for a unit duct length with different air velocities and duct
diameters in common use and corresponds to insulation thickness of li = 30mm and
thermal conductivity λi = 0,035 W/(m2.K).
0,14 D100
0,12 D160
D250
∆t1/(to-tin), [m-1]
0,10
D400
0,08
D630
0,06 D1000
0,04
0,02
0,00
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0
Air velocity, m/s
Figure 6.4. Duct heat gain: the effect of air velocities for temperature change. The
diagram is calculated for a unit length of insulated duct with the lagging
thickness li=30mm and λi = 0,035 W/(m2°C).
As it can be seen from the figures 6.3 and 6.4, despite the decrease of a U-value, the
temperature rise is considerable once the velocity in the duct falls below about 2,5 m/s.
This especially applies to connection ducts in the system, where the air velocities are
relatively low. For example in a D160 duct with the supply air temperature in the
beginning of the duct tin = +15°C and the outside temperature t0 = +22°C, the air
velocity 0,5 m/s (air flow rate 10 l/s) would lead to temperature increase of +0,3°C in
one meter duct length. This occurs in a duct with lagging of 30 mm and λi = 0,035
W/(m2.K).
The temperature change can be decreased with increased lagging thickness. Figure 6.5
presents the effect of lagging for temperature change. It should be noted that a change
to thicker insulation in considerably short connection ducts may not give that big
effect to overall temperature change. For example the temperature decrease is small
when the lagging is increased from 50mm to 75mm (see figure 6.5).
The temperature change in the duct system with cooled supply air increases
logarithmically along the duct length to the level of outside temperature (see figure
6.1) and therefore the total temperature change in the duct is also dependent on a duct
length. The diagrams 6.4 and 6.5 presented the temperature change for 1 meter of duct
length. For longer ducts the temperature change cannot be calculated by multiplying
the temperature change per meter with the duct length, but must be calculated
separately according to equation 6.5
85
0,280
0mm
0,240 10mm
∆T1/(To-Tin), [m-1] 0,200 30mm

50mm
0,160 75mm
0,120
0,080
0,040
0,000
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0
Air velocity, m/s
Figure 6.5. Duct heat gain: the effect of lagging in reducing temperature change.
The diagram is for insulation λi = 0,035 W/(m.K) and duct size 100mm.
6.2. Simulations and measurements with the duct

system in the field
In order to study more in detail the effects of variable air volume flow to supply air
temperature in the duct, simulations were done with a small part of a VAV system in
Case study 2 (Academicum building). The calculations are based on the equations
presented in the previous chapter 6.1. The scheme of the selected system part is shown
in figure 6.6.
conference
room
11
V&min = 91 l/s 13
V&max =430 l/s

tin=+14,5 ºC
D160
D250
3,0m
D400 D400 D400 D400 D400 D315 D315 D315 D315 D315 D250
2,8m
D160
D160
3,0m
D160
D160
D160
D160
D160
D250
D160
D160
D160
2,5m
1 2 3 4 5 6 7 8 9 10 12
Diffusers 1-11 Diffusers 12-13

V&min = 7 l/s V& 7=l/s7 l/s
min
min
V&max =30 l/s V&max =50 l/s
Figure 6.6. The scheme of the simulated air distribution system used for evaluating
the VAV flow effects on temperature change in the duct system. The
supply air temperature tin = +14,5°C, the room temperature for summer
t0 = +24°C, for winter t0 = +22°C.
86
The described system supplies air to 11 different office rooms (devices 1-11) and to
one conference room (devices 12-13). The maximum designed airflow rate for offices
is 30 l/s and for conference rooms 50 l/s per diffuser. The minimum airflow rate is 7
l/s for all devices. All the connection ducts from the main duct have the same length
2,5m and in conference room 3,0m. The distance between connection ducts is 2,8m.
All ducts are insulated in reality with the insulation thickness li = 30mm, λi = 0,035
W/mK. However, in the simulations different lagging thickness will be tested to
evaluate the effect of insulation to the duct system.
The simulations were done for the summer climate conditions with the room
temperature +24°C. The designed supply air temperature from the diffusers is +15ºC.
The effect of different lagging thickness and airflow rates was studied regarding the
temperature change in the system.
The worst cases seem to occur when the devices are running with the minimum
airflow rates. In these conditions the air velocities in the ducts are lower and the
supply air temperature along the duct increases considerably. As a result, a risk of not
maintaining the required room temperature may occur due to decreased cooling
capacity of supply air. This condition most probably occurs with the connection ducts
at the very end of the main duct.
The described situation may not only take place when all the devices are running with
the minimum airflow rates, e.g. when the rooms are empty, but also when the last
room needs the maximum air while the first rooms are running with the minimum
load, e.g. when devices 1-11 are working with minimum air and last devices 12-13
with maximum air. In this situation the air velocity in the main duct can still be
critically low and the supply air temperature increases significantly along the duct.
Figures 6.7-6.10 present the calculated temperature change in the described system
with maximum and minimum load conditions and with different lagging thickness on
the duct system. As it can be seen from the figures, even with maximum airflow rates
the supply air temperature can increase considerably in the system. The increase can
be up to +2,2ºC compared to the initial supply air temperature, depending on the
thickness of lagging. With minimum airflow rates this increase can be up to +9,5ºC
degrees. The most severe conditions are in the case where there is no insulation at all
on the duct system and the supply air temperature rises to the room temperature level
already inside the duct, before supplying to the room (see figure 6.11).
In general, this means that insulating the ducts in VAV system has a great importance
and should be considered for decreasing the heat gains to the system. However, here
the question arises: which ducts are important to insulate and how thick the insulation
should be in order to achieve the required effect?
87
Main duct L = 29m
Diffusers
17,0
Supply air temperature, °C
CD lCD = 0 mm
16,5
16,0 CD lCD =50mm

CD lCD =30mm CD lCD =75mm
15,5
15,0
14,5 Main duct lMD =30mm
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
Figure 6.7 A change in the supply air temperatures along the duct with maximum
airflow rates and different lagging thickness. The dashed line marks the
air temperature at the beginning of each connection duct inside the main
duct. From the figure: CD- connection duct with the lagging thickness
lCD; lMD- lagging thickness of the main duct. The figure corresponds to
the summer condition t0=+24ºC.
Main duct L = 29m
Diffusers
17,0
CD lCD = 0 mm
16,5
16,0
CD lCD =50mm
15,5 CD lCD =30mm
CD lCD =75mm
15,0
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
Figure 6.8 A change in supply air temperatures along the duct in the simulated
system with maximum airflow rates and different lagging thickness.
The dashed line marks the air temperature at the beginning of each
connection duct inside the main duct. From the figure: CD- connection
duct with the lagging thickness lCD; lMD- insulation thickness of the
main duct. The figure corresponds to summer condition t0=+24ºC.
88
Main duct L = 29m
Diffusers
21,0
20,0 CD lCD =30mm
19,0 CD lCD = 0 mm CD lCD =50mm
18,0 CD lCD =75mm
17,0
16,0
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
system with minimum airflow rates and different lagging thickness. The
dashed line marks the air temperature at the beginning of each connection
duct inside the main duct. From the figure: lMD- insulation thickness of the
main duct, lCD-lagging thickness of the connection duct. The figure
corresponds to summer condition t0=+24ºC.
Main duct L = 29m
Diffusers
21,0
20,0
CD lCD =30mm
19,0 CD lCD = 0 mm
CD lCD =50mm
18,0
CD lCD =75mm
17,0
16,0

14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
system with minimum airflow rates and different lagging thickness. The
dashed line marks the air temperature at the beginning of each
connection duct inside the main duct. From the figure: lMD- insulation
thickness of the main duct, lCD-lagging thickness of the connection duct.
The figure corresponds to summer condition t0=+24ºC.
89
Main duct L = 29m
Diffusers
Supply air temperature, °C 26,0
24,0
CD lCD = 0mm
22,0
20,0
18,0
Main duct lMD = 0mm
16,0
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
system with no insulation at the main duct lMD=0mm and no lagging at
the connection ducts lCD=0mm. The figure corresponds to minimum
airflow rates and summer condition t0=+24ºC. The dashed line marks
the air temperature at the beginning of each connection duct inside the
main duct.
Figures 6.7 - 6.10 show that depending on the airflow rates, the supply air can heat up
approx +2,0ºC in the connection duct compared to the air temperature in the main
duct, which presented as a dashed line on these figures. This happens when there is no
insulation on the connection ducts. By insulating the connection ducts this effect can
obviously be decreased, as also shown in figures.
However, it can be also seen that the increase in lagging thickness from 30mm to
75mm gives relatively small effect to overall temperature change in these ducts.
Moreover, since the supply air temperature from the diffusers is directly dependent on
the air temperature at the beginning of the connection duct, controlling the temperature
change in the in the main duct can give even bigger effect to overall temperature
change. For example the supply air temperature from the diffusers will be lower when
the lagging on the main duct is 50mm and connection ducts with 30mm compared to
30/50mm case (see fig. 6.9 and 6.10).
Figures 6.12 and 6.13 present the situation with the supply air temperature when the
first 11 diffusers are running with the minimum airflow rates and the last 2 devices are
working under the full load condition. This situation can occur for example, when all
the office workers from the rooms 1-11 have left their rooms to take part in a meeting
in the conference room.
In this case, as can be see from the figures, it can be a problem to keep the required
supply air temperature even with higher lagging thickness. The supply air temperature
from the end devices will be approx +1,5ºC higher compared to the designed supply
90
air temperature +15ºC. This is due to the fact that air velocity in the main duct is still
relatively low and the supply air temperature in the main duct will increase
considerably before it reaches the last outlets. The jump in the diagram in supply air
temperatures at the end is due to change of air velocity in the duct. With 50 l/s in a
D250 duct the air velocity is 1,0 m/s, whereas with 7 l/s the air velocity is 0,14 m/s.
As one possible solution for solving the problem presented on diagrams 6.12 and 6.13,
would be increasing the air velocity in the main duct by increasing the minimum set
airflow rates from the first 11 diffusers. Figures 6.14 and 6.15 show the results of this
simulation, where the minimum airflow rate from the diffusers 1-11 was set to 10 l/s,
while the last two devices 12-13 are running with full load 50 l/s.
However, according to the calculated numbers, increasing the minimum airflow rates
by 40% gives a very modest temperature decrease, only up to +0,1ºC. Therefore
keeping the minimum airflow rates higher in order to increase the air velocities in the
duct may not give the desired effect.
Another option can be to consider this higher supply air temperature from the last
connection ducts already in the design process and set the required maximum airflow
rates respectively. This would help to assure that the required room temperature would
be met also under these extreme conditions that in reality can occur.
91
Main duct L = 29m
Diffusers
17,5
17,0 CD lCD =30mm
16,5
CD lCD =50mm
16,0
15,5
14,5
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
system with minimum airflow rates from the first 11 diffusers and
maximum airflow rate from the last diffusers 12-13. The dashed line
marks the air temperature at the beginning of each connection duct
inside the main duct. From the figure: lMD- insulation thickness of the
Main duct L = 29m
Diffusers
17,5
17,0
16,5 CD lCD =30mm
16,0
15,5 CD lCD =50mm
15,0
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
system with minimum airflow rates from the first 11 diffusers and
maximum airflow rate from the last diffusers 12-13. The dashed line
marks the air temperature at the beginning of each connection duct
inside the main duct. From the figure: lMD- insulation thickness of the
92
Main duct L = 29m
Diffusers
Supply air temperature, °C 17,5
17,0
16,5 CD lCD =30mm
16,0 CD lCD =50mm

15,5
15,0
Main duct lMD =30mm
14,5
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
Figure 6.14 A change in supply air temperatures along the with minimum airflow
rates V=10 l/s from the first 11 diffusers and maximum airflow rates
V=50 l/s from the last diffusers 12-13. The dashed line marks the air
temperature at the beginning of each connection duct inside the main
duct. From the figure: lMD- insulation thickness of the main duct, lCD-
lagging thickness of the connection duct. The figure corresponds to
summer condition t0=+24ºC.
Main duct L = 29m
Diffusers
17,5
17,0
16,5
CD lCD =30mm
16,0
15,5 CD lCD =50mm
15,0
Main duct lMD =50mm
14,5
14,0
1 2 3 4 5 6 7 8 9 10/11 12/13
Diffuser nr
Figure 6.15 A change in supply air temperatures along the duct with airflow rates
V=10 l/s from the first 11 diffusers and maximum airflow rates V=50 l/s
from the last diffusers 12-13. The dashed line marks the air temperature
at the beginning of each connection duct inside the main duct. From the
figure: lMD- insulation thickness of the main duct, lCD-lagging thickness
of the connection duct. The figure corresponds to summer condition
t0=+24ºC.
93
The simulated air distribution system part described so far, was also tested in the field
in order to see if the temperature change in reality is similar do the simulations done
by modelling. The duct air temperatures were measured in 7 different points in the
system and the results compared with the modelling results. The measurements were
done with 7 different airflow conditions, where the airflow rates from the diffusers
were gradually changed from minimum to maximum (see table 6.1).
Table 6.1.
The tested airflow conditions for the supply air temperature measurements in the
simulated duct system. The airflow rates from 13 diffusers in the test system were
changed gradually from minimum to maximum. The inaccuracy of the flow
measurements is ±2 l/s.
Airflow rates
Diffuser 1 2 3 4 5 6 7 8 9 10 11 12 13
Case 1 5 8 7 9 7 8 6 8 9 6 6 10 10
Case 2 5 7 7 7 6 6 6 7 9 10 9 46 41
Case 3 5 7 7 5 6 5 5 6 7 26 27 41 50
Case 4 5 6 7 5 6 5 5 26 26 26 27 37 44
Case 5 5 6 7 5 6 26 28 29 26 29 28 35 43
Case 6 5 5 7 26 26 27 26 27 22 27 29 33 40
Case 7 26 26 7 27 29 28 27 27 26 28 27 29 38
Unfortunately the initial duct temperature was somewhat higher (+16,3ºC) compared
to the design conditions described in previous figures, which was +14,5ºC. This was
due to the changed settings in the central air-handling unit. However, when comparing
the measured results with the calculated ones, the higher initial temperature was taken
into account.
Figure 6.16 gives the results of the measured air temperatures in these 7 system points
during the measurement period. The measurement points 1, 4, 5 and 7 were inside the
main duct, the points 2, 3, 6 and 8 correspond to supply air temperatures from the
diffusers.
According to the measurement results, the air temperature in the points 6 and 8, which
correspond to the supply air temperatures from the diffusers, decreased considerably
after the airflow rates from the devices were increased. However, despite the increase
in air velocities in the main duct due to the increased airflow rates, the supply air
temperatures in points 2 and 3 did not decrease, unless the airflow rate from those
devices was changed to maximum, which was in case 6 and case 7. The reason for this
is hard to identify. According to the calculations, the temperature in these points
should have decreased gradually with the increased airflow rates.
94
6
1 4 5 7
2 3 8
case 1 case 2 case 3 case 4 case 5 case 6 case 7
19,0
3
18,5
7 2
Temperature, °C
18,0
6
17,5
17,0 4 8
16,5
5
16,0 1
15,5
0 15 30 45 60 75 90 105 120
Time, min
Figure 6.16 Measured supply air temperatures in different points in the simulated
system part in Case study 2 (Academicum, Sahlgrenska Academy) and
with different airflow cases. The lines on the diagram correspond to
measured temperature values in the system points marked in the boxes.
The scheme above the diagram gives the exact location of the point
where the measurement was taken.
Figure 6.17 gives the comparison between the results from the calculations and results
measured in the field. The measured values in the figure are the average over the last
minutes of each case, after the temperature in the measurement point got stable after
the airflow change in the system. In the ideal case, if the both of the results should
match, the points in the figure should be on the diagonal line. It can be seen that with
lower air temperatures there is a better fit between the measurements and a model
compared with higher duct air temperatures.
As mentioned before, the biggest discrepancies between the model and measurements
occurred in the measurement points 2 and 3. It should be also noted that in the
calculations the duct insulation was considered to be l = 30mm and with the thermal
conductivity λi = 0,035 W/(m.K). One of the reasons why the temperature was
considerably higher in these connection ducts may also be not so good insulation of
these ducts. For example in the last connection duct, which corresponds to
measurement points 7 and 8, the differences between the model and the measurement
are smaller (see table 6.2).
95
Calculated temperature in the duct, °C
19,5
cas e 1
19,0 cas e 2
cas e 3
18,5 cas e 4
cas e 5
18,0
cas e 6
17,5 cas e 7
17,0
16,5
16,0
16,0 16,5 17,0 17,5 18,0 18,5 19,0 19,5
Measured temperature in the duct, °C
Figure 6.17 A comparison between the calculated and measured results from the
supply air temperature evaluation in the duct system. In the ideal case,
the points in the diagram should follow the diagonal line.
Table 6.2.
A comparison between the calculated and measured results from the supply air
temperature evaluation in the duct system. The measurement points 2, 3, 6 and 8
correspond to supply air temperatures from the diffusers.
Measured point 1 2 3 4 5 6 7 8
Case 1 V l/s 99 8 7 70 49 6 20 10
temp measured 16,4 17,6 18,6 17,4 17,3 18,6 17,8 18,5
temp model 16,5 17,1 17,6 17,0 17,6 18,7 18,5 19,1
difference -0,1 0,5 1,0 0,4 -0,3 -0,1 -0,7 -0,6
Case 2 V l/s 166,0 7,0 6,0 140,0 122,0 9,0 87,0 41,0
temp measured 16,4 17,7 18,7 17,2 16,9 18,7 17,2 17,6
temp model 16,4 17,1 17,4 16,8 17,1 17,8 17,4 17,6
difference 0,0 0,6 1,3 0,4 -0,2 0,9 -0,2 0,0
Case 3 V l/s 197,0 7,0 6,0 173,0 157,0 27,0 91,0 50,0
temp measured 16,3 17,8 18,8 17,1 17,5 17,4 16,9 17,3
temp model 16,3 17,0 17,3 16,6 16,9 17,2 17,1 17,3
difference 0,0 0,8 1,5 0,5 0,6 0,2 -0,2 0,0
Case 4 V l/s 225,0 6,0 6,0 202,0 186,0 27,0 81,0 44,0
temp measured 16,3 17,8 18,9 17,0 16,7 17,3 16,8 17,2
temp model 16,3 17,0 17,3 16,6 16,8 17,2 17,1 17,3
difference 0,0 0,8 1,6 0,4 -0,1 0,1 -0,3 -0,1
Case 5 V l/s 273,0 6,0 6,0 250,0 190,0 28,0 78,0 43,0
temp measured 16,3 17,9 18,9 17,0 16,6 17,3 16,8 17,1
temp model 16,3 17,0 17,2 16,5 16,7 17,1 17,0 17,2
difference 0,0 0,9 1,7 0,5 -0,1 0,2 -0,2 -0,1
Case 6 V l/s 300,0 5,0 26,0 257,0 178,0 29,0 73,0 40,0
temp measured 16,2 17,9 17,1 16,6 16,6 17,3 16,8 17,1
temp model 16,2 17,0 16,6 16,4 16,6 17,0 16,9 17,1
difference 0,0 0,9 0,5 0,2 0,0 0,3 -0,1 0,0
Case 7 V l/s 345,0 26,0 29,0 259,0 175,0 27,0 73,0 38,0
temp measured 16,2 16,9 16,9 16,5 16,5 17,3 16,8 17,1
temp model 16,2 16,5 16,6 16,4 16,6 17,0 16,9 17,2
difference 0,0 0,4 0,3 0,1 -0,1 0,3 -0,1 -0,1
96
In general, as can be also seen from the figure 6.16, the duct air temperature is
decreasing considerably after the airflow is increased. In a building in operation the
airflow rates are controlled according to the demand of the rooms/zones and as soon as
the temperature starts to rise in a room, more air will be supplied.
The analysed cases described in figures 6.9 – 6.14 occur mainly when most of the
rooms are empty and in these situations the supply air temperature will not have a
great importance. Nevertheless, since the system is controlled by the room
temperature, higher supply air temperatures and due to that poorer cooling capacity of
the supply air can lead to the increase of supply airflow rates in order to maintain the
required room temperature. This is however, a waste of energy and therefore methods
for decreasing the heat gains to the duct system should really be considered already in
the design process.
To conclude, the methods that can be applied in order to maintain the required supply
air temperatures in a VAV duct system are following:
• Insulating all the ducts in the system and using bigger insulation thickness on
the main ducts (50mm can be recommended). As was shown by the
simulations, the bigger effect for decreasing heat gains to the duct system was
achieved when the lagging on the main ducts was increased instead of
increasing the lagging on the connection ducts.
• Considering higher supply air temperatures from the last connection ducts
already in the design process and determining the required airflow rates to
these rooms according to this temperature increase in the system. This can
assure that the required room temperatures are achieved also with these severe
conditions that can occur if the airflow rates vary in the system……
97
98
7 CONCLUSIONS AND FUTURE RESEARCH
7.1 Conclusions
The aim of this work was to clarify prerequisites for well functioning VAV system in
office buildings.
Firstly the fundamental demands were set for the system and its components in general,
with the focus on indoor climate. To achieve a good indoor climate is the main task of
HVAC systems. This should be done with lowest possible need of energy.
Secondly essential technical properties for fulfilling the indoor climate demands were
identified. When doing this one goal was to look for an uncomplicated system solution,
which should be possible to implement both in existing buildings and in new ones. One
criterion of an uncomplicated system is that the number of controlling components is
minimized. If the supply air diffusers can manage quite a high pressure drop, it may be
possible to obtain a good over all function without controlling dampers in the duct
system.
Furthermore, in order to fulfil the indoor climate aspects, the diffuser should have a
stable air movement pattern, which must be independent of the supply airflow rate. E.g.,
neither at high air flow rates, nor low airflow rates, should there be any risk for cold
supply air dropping into occupied zone. The diffuser should also be able to tolerate a
high pressure drop without disturbing generation of noise. A pressure drop of at least
100 Pa over the diffuser should be managed.
In order to test such an uncomplicated system configuration, a type of supply air device,
that seemed to fulfil the above demands, was installed in two buildings, one existing
and one new. In the existing building case the old CAV system was changed to VAV. In
the new building it was a direct VAV design.
The function of both plants was monitored and tested. The results show that the indoor
climate demands are essentially fulfilled.
• No risk of draught was indicated when the pressure independent VAV supply air
diffusers were operating with the maximum flow conditions and with +15°C
supply air temperature.
• The noise levels were acceptable. The sound pressure level in the measured
rooms was lower than 30 dB(A) even when the pressure drop over the device
was around 100 Pa.
This verifies that the system configuration, without controlling dampers, functions as
expected from thermal comfort point of view, both when applied in an existing system
and when installed in a new building.
In order to determine the thermal comfort properties more in detail the VAV supply air
diffuser was also tested under laboratory conditions. In a full scale test room thermal
99
comfort parameters were studied under different flow conditions and different heat
loads. All tests were carried through with the supply air temperature about +15oC.
The results from the laboratory study indicated following:
• At medium airflow (25 l/s) and minimum airflow (10 l/s) airflow the air
movements and the draught levels in the room did not exceed the required levels
stated by comfort standards and regulations.
• Even at the lowest airflow rate, no risk of “air-dumping” was indicated.
• At the highest airflow rate (50 l/s) a marginal draught risk was registered in a
few measured points (in 10% of the measured points).
• All these points with registered draught risk were situated on the empty side of
the test room, opposite the workplace. No draught risk was observed in the
normal working zone. Since the air motion in the room is complex, it is hard to
make any definite conclusions for the causes of the draught risk in these
specified room points and more research should be conducted in this area.
• There is a direct relation between air velocity and turbulence intensity.

Moreover, the points with higher turbulence intensities were situated close to the
floor, measured at ankle level. Similar findings are accounted for different air
diffusers by other researches [12]. However in the present case, this higher air
turbulence on the ankle level may have partly been caused by a floor
temperature, which was some degrees lower than the room temperature.
In general it can be concluded that the air supplied by a VAV diffuser should have a
high momentum within the whole airflow range. This can ensure high induction
properties and a good mixing of cold supply air into room air at all airflow conditions.
From the measurements done in the laboratory and in the field with one type of a
variable VAV device, the risk of draught was considerably low despite the high airflow
rates and low supply air temperatures.
The field monitoring of energy consumption and airflow rates showed that the tested
VAV systems are energy efficient. Due to the low supply air temperatures, about
+13.5°C to +14°C from the central air handling unit, the heat recovery system accounts
for almost all the heating of air needed. There is almost no need for additional heating
with the heating coil.
Due to the low supply air temperature, the airflow rate regulation versus heat load is
effective. This contributes to a low average airflow rate, and therewith the energy need
for air distribution becomes relatively low.
The influence of a variable air volume flow on duct heat gains was also analysed and
mathematically evaluated in this current study. The cold air streams in ducts are
influenced by heat gains due to the temperature difference between the supply air in the
duct and the room temperature.
100
The worst cases occur when the devices are running with minimum airflow rates. In
these conditions the air velocities in the ducts are the lowest and consequently the
temperature of supply air increases considerably along the duct. As a result, a risk of not
maintaining the required room temperature may occur due to decreased cooling capacity
of the supply air. This is most likely to occur in the rooms at the very end of the duct
line.
From the results of the simulations done with the test system, some methods were
recommended that could be applied in order to maintain the required supply air
temperatures in a VAV duct system. The focus here was on lagging the main ducts with
bigger insulation thickness, since this can give a much bigger effect than increasing the
lagging thickness on connection ducts.
Nevertheless, in some system parts, especially at the very end of the duct lines, the
supply air temperature will be heated up the most and the use of bigger insulation
thickness may still not assure the design supply air temperature. One of the options here
could be considering this higher supply air temperature already in the design process
and determining the required airflow rates to these rooms according to this temperature
increase in the system. This can assure that the required room temperatures are achieved
also with these severe conditions that can occur in the system with varying airflow rates.
7.2 Future studies

One of the results from the case studies in this current project showed that even though
the systems generally had a low use of energy, the measured systems hardly ever
reached their designed values of airflow rates. In two air-handling systems measured the
maximum airflow in practice was less than the design value. This raises some questions.
What is the suitable diversity factor to use in VAV system design in office buildings?
Also, when decreasing the size of ducts and air handling unit due to the diversity factor,
how does this affect the energy use of the air conditioning system?
Nevertheless, selecting the right diversity factor is an issue of itself and it depends to a
great extent on the occupancy levels in rooms. The occupancy is dependent on the
activity in office rooms and can vary widely between office rooms in a university
building and in industrial buildings. A recent Swedish study [36] of three different office
buildings measured the overall average of occupancy to be 46% in the daytime. In a
university building the average day time occupancy rate was approx. 33% and in
municipal office the rate it was approx. 54%. An attempt at modelling [64] the occupancy
rate reported a figure of 78% during the working hours in an office building.
In both studies the chosen methodology included using occupancy sensors to register
whether the rooms are occupied or not. The main problem with this methodology is that
it does not consider the number of people in rooms. This means that people can move
from one room to another and never leave the building, leaving the occupancy of the
office building the same. There is a need of further research to determine occupancy
rates and suitable diversity factors for VAV system design in office buildings. These
numbers can help to optimise the size of the components in the system regarding
investment and running cost.
101
102
8 REFERENCES
1. Abel, E., Jagemar, L. and Widén, P., 1997. Energiteknik. Literatur till kursen
Termodynamik och Värmelära. Chalmers University of Tehcnology. Göteborg.
2. Ardehali, M.M. and Smith, T.F., 1996. Evaluation of variable volume and
temperature HVAC system for commercial and residential buildings. Energy
Conversation Management, vol. 37, nr. 9, p. 1469-1479.
3. ASHRAE, 1997. ASHARE Handbook, Fundamentals. American Society of

Heating, Refrigerating and Air-Conditioning Engineers Inc. Atlanta, USA.
4. ASHRAE, 2000. ASHRAE Handbook: HVAC Systems and Equipment. American

Society of Heating, Refrigerating and Air-Conditioning Engineers Inc. Atlanta,
USA.
5. ANSI/ASHRAE Standard 55-2004, 2004, Thermal Environmental Conditions for

Human Occupancy, American Society of Heating, Refrigerating and Air-
Conditioning Engineers, Inc. USA.
6. Bengtsson, M., 2003. The users perception of the indoor environment: A field
survey in a modern office building. Chalmers University of Technology.
Gothenburg, Sweden.
7. Boverket, 1995. Eleffektivitet i byggnader- Handbok The Swedish National Board

of Housing, Building and Planning. Karlskorna, Sweden.
8. BFS 1993:57, 2002, Boverkets Byggregler, BBR (in Swedish), Boverket, The
Swedish National Board of Housing, Building and Planning. Karlskorna, Sweden.
9. Brüel&Kjær, 1996, Thermal comfort. , INNOVA. Naerum, Denmark.
10. Cappellin, T.E., 1997. VAV Systems- What makes them succeed? What makes
them fail? ASHRAE Transactions, vol. 103, nr. 2, p. 814-822.
11. CEN Report CR1752, 1998, Ventilation for buildings- Design criteria for the
indoor environment, European Committee for Standardization.
12. Chow, W.K., Wong, L.T., Chan, K.T. and Yiu, J.M.K., 1994. Experimental
studies on the airflow characteristics of air-conditioned spaces. ASHRAE
Transactions, vol. 100, nr. 1, p. 256-263.
13. CIT Energy Management AB, 2003, BV2. Software for Simulation and Calculation
of Building's Energy Requirement, Gothenburg, Sweden.
14. Daryanani, S., McKay, W., N.Shataloff and H.Straub, 1966. VAV Variable air
volume air conditioning. Air conditioning, heating and ventilating, vol. 63, nr.
March, p. 56-78.
103
15. Day, T.L., 1974. VAV air distribution. ASHRAE Journal, vol. 16, nr. 4, p. 36-40.
16. Eck, B., 1972. Ventilatoren. Springer-Verlag. Berlin.
17. Statens energimyndighet, 2006, Förbättrad energistatistik för lokaler - "Stegvis

STIL" Rapport för år 1. Inventeringar av kontor och förvaltningsbyggnader,
Statens energimyndighet. Sweden.
18. Etheridge, D. and Sandberg, M., 1996. Building ventilation. Theory and
Measurement. Chichester: Wiley.
19. F.Engdahl and A.Svensson, 2003. Pressure controlled variable air volume
systems. Energy and Buildings, vol. 35, p. 1161-1172.
20. F.Engdahl and Johansson, D., 2004. Optimal supply air temperature with respect
to energy use in a variable air volume system. Energy and Buildings, vol. 36, p.
205-218.
21. Fahlén, P., 1996. Frosting and Defrosting of Air Coils - A literature Survey. SP
Swedish National Testing and Research Institute.
22. Fanger, P.O., 1982. Thermal Comfort- Analysis and Applications in

Environmental Engineering. Florida, USA.
23. Fanger, P.O., 2000. Provide good air quality for people and improve their
productivity. Proceedings of the Seventh International Conference of Air
Distribution in Rooms, United Kingdom.
24. Fanger, P.O., Melikov, A.K., Hanzawa, H. and Ring, J., 1988. Air turbulence and
sensation of draught. Energy and Buildings, vol. 12, p. 21-39.
25. Filardo, M.J., 1993. Outdoor air: How much is enough? ASHRAE Journal, vol.
35, nr. 1, p. 34,36-38.
26. Hagentoft, C.-E., 2001. Introduction to Building Physics. Studentlitteratur. Lund,

Sweden.
27. Mark Hydeman, Steve Taylor and Jeff Stein, 2003, Advanced Variable Air
Volume System Design Guide, California Energy Commission. California.
28. Inoue, U. and Matsumoto, T., 1979. A Study on Energy Savings with Variable Air
Volume Systems by Simulation and Field Measurement. Energy and Buildings,
vol. 2, p. 27-36.
29. Int-Hout, D. and Berger, P., 1984. What´s really wrong with VAV Systems.
ASHRAE Journal, vol. 26, nr. December, p. 36-38.
30. ISO7730, 1995, Moderate Thermal Environments- Determination pf the PMV and
PPD indices and specification of the conditions for thermal comfort, ISO
(International Organization for Standardization).
104
31. ISO 10551, 1995, Ergonomics of the thermal environment- assessment of the
influence of the thermal environment using subjective judgement scales., ISO
(International Organization for Standardization).
32. ISO, 1993, Guide to the Expression of Uncertainty in Measurement, ISO

(International Organization for Standardization). Switzerland.
33. Jagemar, L., 1996. Design of energy efficient buildings - Applied on HVAC
Systems in Commercial Buildings. PhD thesis, Building Services Engineering,
Chalmers University of Technology, Göteborg.
34. Jagemar L. , 2005, Mätning och utvärdering av VAV-system på del av EDIT-huset

(O7:20) på Chalmers Campus Johanneberg i Göteborg, Slutrapport, revision 1,
CIT Energy Management AB. Göteborg.
35. Jagemar L. , 2005, Utvärdering av energi-och luftbehandlingssystemen på

Academicum och Gamla Anatomihöghuset, Medicinareberget i Göteborg,
Slutrapport / revision 1, CIT Energy Management AB. Göteborg.
36. Johansson, D., 2005. Modelling Life Cycle Cost for Indoor Climate Systems. PhD
thesis, Building Physics, Lund Institute of Technology, Lund.
37. Johnson, G.A., 1985. From constant air to variable. ASHRAE Journal, vol. 27, nr.
1, p. 106-111, 114.
38. Jones, W.P., 2001. Air Conditioning Engineering. Edward Arnold. London.
39. Kloostra, L., 1979. VAV systems saves 38% of energy use. Heating, piping, air
conditioning, vol. 51, nr. 12, p. 61-63.
40. Lagoudi, A., Loizidou, M., Santamouris, M. and Asimakopoulos, D., 1996.
Symptoms experienced, environmental factors and energy consumption in office
buildings. Energy and Buildings, vol. 24, p. 237-243.
41. Li, Z.H., Zhang, J.S., Zhivov, A.M. and Christianson, L.L., 1993. Characteristics
of diffuser air jets and airflow in the occupied regions of mechanically ventilated
rooms- A literature review. ASHRAE Transactions, vol. 99, nr. n1, p. 1119-1127.
42. Linder, R. and Dorgan, C.B., 1997. VAV Systems Work Despite Some Design and
Application Problems. ASHARE Transactions, vol. 103, nr. p2, p. 807-813.
43. Maki, K.S., Li, Z., Chamberlin, G.A. and L.L.Christianson, 1997. VAV Systems
Performance- Field Characerization of Flow, System Diagnosis Tools, and
Operational Design Implications. ASHRAE Transactions, vol. 103, nr. 2, p. 830-
842.
44. Miljöförvaltningen, 2005, Luftkvalitet i Göteborgsområdet, Miljöförvaltningen-

Göteborg. www.miljo.goteborg.se/luftnet. Göteborg, Sweden.
105
45. Mull, T.E., 2004. Energy conservation measures for air distribution and HVAC
systems. Plant Engineering, vol. 58, nr. 10, p. 60-64.
46. Mysen, M., Rydock, I.P. and Tjelflaat, P.O., 2003. Demand controlled ventilation
for office cubicles- can it be profitable? Energy and Buildings, vol. 35, p. 657-
662.
47. Nilson, A. and Uppström, R., 1996. ENEU 94 K. Guidelines for the Procurement
of Energy-Efficient Equipment in Public Services. Bengt Dahlgren AB.
Stockholm.
48. Nilson, A., Uppström, R. and Hjalmarsson, C., 1997. Energy Efficiency in Office
Buildings. Lessons from Swedish Projects., Swedish Council for Building
Research. Stockholm.
49. Nilsson, P.E., 2003. Achieving the Desired Indoor Climate. Energy Efficiency
Aspects of System Design. red., The Commtech Group. Studentlitteratur.
50. Norford, L.K., Rabl, A. and Socolow, R.H., 1986. Control of supply air
temperature and outdoor airflow and its effect on energy use in a variable air
volume system. ASHRAE Transactions, vol. 92, p. 30-45.
51. Oseland, N.A., 1995. Predicted and reported thermal sensation in climate
chambers, offices and homes. Energy and Buildings, vol. 23, p. 105-115.
52. R1, 2000, Classified Indoor Climate Systems - Guiding principles and
Specifications VVS Tekniska Föreningen/Förlags AB. Ödeshög, Sweden.
53. Rickelton, D. and Becker, H.P., 1972. Variable Air Volume. ASHRAE Journal,
vol. 9, nr. 14, p. 31-55.
54. Schiller, G.E., 1990. A comparison of measured and predicted comfort in office
buildings. ASHRAE Transactions, vol. 96, nr. 1, p. 609-622.
55. Shepherd, K., 1999. VAV Air Conditioning Systems. Blackwell Science Ltd.
United Kingdom.
56. Skoog, J., 2004. Indoor Environment and User Perception. A Field Study in a
Hospital Ward. Licentiate thesis, Building Services Engineering, Chalmers
University of Technology, Gothenburg, Sweden.
57. Svensk Standard SS 02 52 68, 2000, Byggakustik - Ljudklassning av utrymmen i

byggnader - Vårdlokaler, undervisninglokaler, dag- och fritidshem, kontor och
hotell, Standardiseringen i Sverige.
58. Stoops, J.L., 2000. The thermal environment and occupant thermal perceptions in
office buildings- An evaluation of sampled data from five European countries.,
Licentiate thesis, Building Services Engineering, Chalmers University of
Technology, Gothenburg, Sweden.
106
59. Stoops, J.L., 2001. The physical environment and occupant thermal perceptions in
office buildings. An evaluation of sampled data from five european countries.,
Ph.D Thesis, Building Services Engineering, Chalmers University of Technology,
Gothenburg, Sweden.
60. Straub, H.E., 1970. Room Air Distribution with a Variable Volume System.
ASHRAE Journal, vol. 12, nr. April, p. 52-58.
61. Tamblyn, R.T., 1983. Beating the blahs for VAV. ASHRAE Journal, vol. 25, nr.
September, p. 42-45.
62. Tozour, D.O., 1986. Energy retrofit makes old building young again. Heating,
piping, air conditioning, vol. 58, nr. 3, p. 115-117.
63. M3003, 1997, The Expression of Uncertainty and Confidence in Measurement,

UKAS (United Kigdom Accreditation Service). United Kingdom.
64. Wang, D., Federspiel, C.C. and Rubinstein, F., 2005. Modeling Occupancy in
Single Person Offices. Energy and Buildings, vol. 37, nr. 2, February, p. 121-126.
65. Wargocki, P., Wyon, D.P., Sundell, J., Clausen, G. and Fanger, P.O., 2000. The
Effects of outdoor air supply rate in an office on precieved air quality, sick
building syndrome (SBS) symptoms and productivity. Indoor Air, vol. 10, p. 222-
236.
66. Wendes, H., 1989. Supply outlets for VAV systems. Heating, piping, air
conditioning, vol. 61, nr. n2, p. 67-71.
67. Zheng, G.R. and Zaheer-Uddin, M., 1996. Optimization of thermal process in a
variable air volume system. Energy vol. 21, nr. 5, p. 407-420.
68. Åström, J., 2004. Measurement of Loss Components in two Induction Machines
and three types of Frequency Converters. master thesis, The Department of
Electric Power Engineering, Chalmers University of Technology, Gothenburg,
Sweden.
107
108
APPENDICES
APPENDIX A
Measurement instruments and measurement uncertainties
APPENDIX B
Results from the laboratory measurements
APPENDIX C
Questionnaire
The questionnaire presented here is from the summer study. The winter study
questionnaire is identical except for the title page.
109
110
A Measurement instruments and
measurement uncertainties
A.1 Field measurements
Two different measurement instruments were used for the measuring of thermal comfort
parameters in the case studies. A preliminary requirement for selecting an instrument for
these measurements was to have an omni directional air velocity transducer that is able to
measure air velocity down to 0,05 m/s and fluctuations up to 2 Hz [9]. This is important for
evaluating draught rate in the room.
A “Brüel & Kjaer Model 1213 – Thermal Climate Analyzer” was used for the Case study
1 for the measuring of room temperatures and air velocities in selected office rooms. The
instrument also calculates a standard deviation of an air velocity in a measured room
point.
A SWEMA 300 measurement instrument with SWA 01 comfort probe was used for the
same kind of measurements in Case study 2. The instrument measures and displays the
following parameters:
• room temperature
• average air velocity
• minimum air velocity
• maximum air velocity
• standard deviation of air velocity
• Draught Rate, DR
The room temperature and average air velocity values are measured and averaged over a
three-minute period.
The characteristics of both instruments are given in table A.1.
Table A.1.
Information about the instruments used for thermal comfort measurements
Instrument Measuring Measuring Inaccuracy Inaccuracy
Range, [°C] Range, [m/s] [°C] [m/s]
B&K 1213 -20°C…50°C 0,05…1 m/s ± 0,2 °C ± 0,02 m/s
Thermal climate analyser
SWA 01+ 10°C…40°C 0…1,0 m/s ± 0,3 °C ± 0,02 m/s

SWEMA 300logger
The supply air temperature in air ducts before the supply air devices was measured with
probe SWA 31 connected to SWEMA 300 logger, which measures the temperatures in the
range of –20…+80°C with the inaccuracy of ± 0,3 °C. The given inaccuracy includes a
probe together with any calibrated SWEMA Air 300 logger.
The air velocities from the variable supply air diffusers were measured with a
measurement equipment installed into the device and the results were read with a palm
111
computer. According to the data from the manufacturer, the inaccuracy of the flow
measurement sensors of the supply air diffuser is ± 2 l/s.
The noise measurements were carried out with a “Brüel & Kjær” sound analyser, type
2260 Investigator including application BZ7206 version 2.1. The sound pressure levels
were measured for different frequencies divided in thirds of the octave band. The
measurement range was set to 9,7-89,7 dB. An “A” frequency weighting was used. Before
starting up the measurements the instrument was externally calibrated in the field with the
manufacture’s sound level calibrator.
The pressure in the duct before the air supply outlets was measured with the Prandtl-tube,
which has the inaccuracy of reading the values from the scale approximately half of the
value between the ticks on the scale. Thus the inaccuracy is ± 0,5 Pa.
For the heat transfer analysis when evaluating the heat gains to the supply air, the air
temperatures in different points in the ducts were measured with Pt-100 temperature
sensors, which together with logger have the inaccuracy of ± 0,061 °C.
In both case studies, the airflow and the energy measurements in the central air-handling
units were done by a separate company “Mätcentralen”. The detailed analysis of the
instruments used in these measurements is not included in this work. However, all of the
values lower than the minimum airflow rate in the system were considered to be part of
the measurement error and the corresponding values were not accounted in the analysis.
The test room temperature was also monitored together with the laboratory hall
temperature with Pt-100 sensors. The same type of sensors was used for measuring
operative temperature in the test room and the supply air temperature in the duct. All the
sensors, which were connected to the logger and PC software, were calibrated using
Mercury thermometer, which has the inaccuracy of ± 0,05 °C.
A.2 Laboratory measurements

Three SWEMA 300 measurement instruments with SWA 01 and SWA 03 comfort probes
were used for measuring thermal comfort parameters in the test room. The characteristics
of the sensors are given in table A.1 above. The draught probe SWA 01 is an older type of
this type of a sensor, but the measurement characteristics are the same as SWA 03. The
time constant of the sensors is 0,25 seconds.
The test room temperature and the laboratory hall temperature were measured with Pt-100
sensors. The same type of sensors was used for measuring plane radiant temperatures in
the test room and the supply air temperature in the duct. For estimating the operative
temperature a plane radiant temperature for six different directions was measured and the
operative temperature was calculated according to equations 4.1 and 4.2 [9, 30]. given in
chapter 4.2.
All the sensors, which were connected to the logger and PC software, were calibrated
using Mercury thermometer, which has the inaccuracy of ± 0,05 °C.
112
The airflow rates in the system were measured with flow measuring devices installed in
the ducts. The lower airflow rates 10 and 25 l/s were measured with “Fläkt Woods” IRIS
damper and the higher airflow rates up to 50 l/s were measured with an adjustable circular
measuring damper (from “Swegon” type CRMc). According to the manufacturers data the
method error for the IRIS type of a damper is ± 7% and for the other damper it is ± 5%.
The differential pressure for both measuring devices was measured with electronic
pressure sensors and the corresponding airflow rates were calculated according to the
equation A.1:
V& = K ∆p m (A.1)
Where,
V& air flow rate, l/s
K factor which depends on the airflow rate and on the device setting
∆pm measured pressure difference in the measuring device, Pa
The K factors for different device obstructions that were used for measuring different
airflow rates are given in table 2.
Table A.2.
The K-factors for flow measurement dampers.
Damper Airflow rate, l/s K-factor
IRIS 10 1,6
24 4,8
CRMc 50 15,8
The electronic pressure sensor was also used for the measuring of the pressure before the
supply air device in the room. The inaccuracy of the pressure transducers is ± 1%.
All the sensors (temperature and pressure sensors) were connected to a PC-logger 3100
and monitored from the computer with EasyView software version 5.5 Pro.
A.3 Measurement uncertainties

The components of measurement uncertainty can be categorized, according to the method
used to evaluate them, in two groups [32]:
• Type A: those that are evaluated by statistical methods
• Type B: those that are evaluated by other means
All the uncertainty components are modelled by probability distribution quantified by

variances or standard deviations. The estimated standard deviation is defined as a standard
uncertainty and represented by u(xi).
A type A evaluation is used to obtain a value for the repeatability or randomness of a

measurement process. It is represented by a statistically estimated standard deviation of a
sample mean of n independent repeated values for a quantity q. For such a component the
standard uncertainty is u ( xi ) = s(q ) and can be expressed by equation A.2 [63]:
113
1
⎡ 1 n ⎤2
u ( xi ) = s ( z ) = ⎢ ∑ (z j − z)2 ⎥ (A.2)
⎣ n(n − 1) j =1 ⎦
Where ,
u ( xi ) standard uncertainty associated with the estimated value of each input quantity xi.
The input quantities contribute to the estimated value of the output quantity y in
function form as follows: y = f(x1, x2,…, xN).
s ( z ) standard deviation of a sample mean of n independent repeated values for a
quantity z
zj jth repeated observation of randomly varying quantity z
n number of observations in a sample
z arithmetic mean of n repeated observations of randomly varying quantity q
The uncertainty components obtained by a type B evaluation are systematic components

of uncertainty. They account for errors that remain constant while the measurements are
made. These are for example uncertainties associated with measurement instruments,
measurement methods, etc.
Similarly to Type A uncertainty components the standard uncertainty u(xi) for Type B
evaluation is equal to an approximation to the corresponding standard deviation obtained
from an assumed probability distribution. This assumed probability distribution is based
on all the available information, which may include previous measurement data,
manufacturer’s specifications, data provided in calibration reports, etc.
Once the standard uncertainties u(xi) of the input quantities xi have been derived from
both Type A and Type B evaluations, the standard uncertainty of the output quantity
y=f(x1, x2, …,xN), can be calculated as follows [63] :
1
⎡N ⎤2
u c ( y ) = ⎢∑ ci2 ⋅ u 2 ( xi )⎥ (A.3)
⎣ i =1 ⎦
Where,
ci sensitivity coefficient. It is equal to a partial derivate with respect to input quantity
xi of the functional relationship f between the measurement output and the input
quantities.
N number of input estimates xi on which the measurement output depends
In order to provide a confidence interval for the output estimate y, an expanded

uncertainty should be calculated. It is obtained by multiplying the combined standard
uncertainty by a coverage factor k:
ue = k ⋅ uc ( y) (A.4)
Where,
ue expanded uncertainty of output estimate y that provides a confidence interval Y =
y ± ue
114
k coverage factor. A value k = 2 gives a level of confidence of approx. 95%
Tables A.3 to A.5 give the measurement uncertainty components and expanded
uncertainties for calibration of the measurement chain of temperature and airflow
measurements in the laboratory hall. The combined standard uncertainties and expanded
uncertainties are calculated according to equations A.3 and A.4. The description of the
measurement instruments and their associated inaccuracies were discussed in previous
chapter A.2 in this APPENDIX A.
Table A.3
Measurement uncertainty components and expanded uncertainties for temperature
measurements with Pt100 sensors.
Symbol Source of uncertainty Value Probability Divisor ci ui(ti)
°C distribution °C
ut1 Calibration with the reference 0,05 rectangular √3 1.0 0.03
thermometer
ut2 Resolution of the logger 0,01 rectangular √3 1.0 0.006
uc(t) Combined standard uncertainty normal 0.03
ue Expanded uncertainty normal (k=2) 0.06
Table A.4
Measurement uncertainty components and expanded uncertainties for differential pressure
measurements for the airflow measurement device IRIS.
% distribution %
udp1 Method uncertainty of IRIS 7 rectangular √3 1.0 4
damper
udp2 Inaccuracy of electronic 1 normal 1.0 1.0 1
pressure sensor
udp3 Uncertainty from the logger 0.05 normal 1.0 1.0 0.05
Table A.5
Measurement uncertainty components and expanded uncertainties for differential pressure
measurements for the airflow measurement device CRM160.
% distribution %
udp1 Method uncertainty of flow 5 rectangular √3 1.0 2.9
measuring device CRM
udp2 Inaccuracy of electronic 1 normal 1.0 1.0 1
pressure sensor
udp3 Uncertainty from the logger 0.05 normal 1.0 1.0 0.05
The expanded uncertainty for airflow measurement with the IRIS damper is ± 4,1%. This
expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor
k= 2, providing a level of confidence of approx. 95%. For the airflow measurement
device CRM160 the expanded uncertainty is ± 3,1%.
115
The expanded uncertainty for operative temperature measurement with six Pt-100
temperature sensors is ± 0,15°C (k = 2).
The temperatures monitored with Pt-100 sensors consisted of a series of observations

registered each second. Each sample taken over the measurement period during the
steady-state conditions of the environment can be considered as a repeated measurement-
a replicate. The expanded uncertainties for these sets of temperature measurements, which
consider both the uncertainty of repeatability of the measurement and the uncertainty of
the calibration of the measurement chain, can be calculated according to eq. A.3 and A.4.
The results are given in the tables B.1 and B.2 in APPENDIX B.
Table A.6 gives the measurement uncertainty components and expanded uncertainty for
calibration of temperature measurement in different room points in the test room. Here
besides the inaccuracy of the instrument, also the uncertainty of positioning the sensor has
been considered. The combined standard uncertainty and expanded uncertainty is
calculated according to equations A.3 and A.4. The description of the measurement
instruments SWEMA 300 with SWA 01 and SWA 03 comfort probes and their associated
inaccuracies were given in previous chapter A.1. The probability distribution for the
measurement uncertainty associated with measurement instrument can be assumed to be
rectangular.
Table A.6
Measurement uncertainty components and expanded uncertainties for temperature
measurements with SWEMA 300 together with SWA01/03 probe.
°C distribution °C
ut1 Inaccuracy of the instrument 0,3 rectangular √3 1.0 0.2
ut2 Method uncertainty 0,05 rectangular √3 1.0 0.03
The thermal comfort measurements conducted in the test room in the laboratory were
done in three replicates for each measurement case. This means that the measurements
were conducted three times with the same environmental and airflow conditions. The
results presented in the tables B.1 and B.2 are average values over these three replicates.
The expanded uncertainty for these measurement sets can be calculated according to
equations A.3 and A.4. It considers both the uncertainty of repeatability of the
measurement and the uncertainty of the calibration of the measurement instrument.
The combined measurement uncertainties for the field measurements, which are mainly
the uncertainties associated with the measurement instruments, were discussed in detail in
previous chapter A.1.
116
B Results from the laboratory measurements
Table B.1.
Measured velocity and temperature values and calculated average draught rating DR for 3
different airflow cases and WITHOUT SUSPENDED CEILING mounting case. The
values in the tables are average values of three replicates made in the same conditions.
Point x.1 is 0,1m above floor, point x.2 is 0,6m above floor and point x.3 is 1,1m above
floor. The uncertainties given for the average operative temperature top, average supply air
temperature tsupply and for the average airflow rate V& are expanded uncertainties
(coverage factor k = 2)
Case 1 Case 2 Case 3
top= 22,6 ± 0,3 °C top= 22,4 ± 0,3 top= 22,4 ± 0,4
tsupply=15,0 ±0,1°C tsupply=15,1± 0,2°C tsupply=15,1 ± 0,2°C
V& = 10,1 ± 0,5 l/s V& = 26,4 ± 1,8 l/s V& = 49,8 ± 4,8 l/s
Poin Temp v DR Temp v DR Temp v DR

t °C m/s % °C m/s % °C m/s %
1.1 21,7 0,03 0,0 21,6 0,05 2,1 21,6 0,06 3,2
1.2 22,2 0,05 0,8 21,9 0,08 4,9 21,7 0,14 11,3
1.3 22,3 0,05 1,0 22,0 0,09 7,0 21,9 0,13 12,0
2.1 21,7 0,05 2,5 21,6 0,11 9,7 21,5 0,09 8,7
2.2 22,1 0,04 0,0 21,8 0,09 7,3 21,7 0,13 12,3
2.3 22,2 0,05 0,9 21,9 0,08 6,0 21,7 0,13 12,5
3.1 21,8 0,04 1,1 21,6 0,09 7,6 21,4 0,13 13,5
3.2 22,1 0,06 2,0 21,8 0,07 4,2 21,6 0,12 11,4
3.3 22,3 0,05 1,7 21,9 0,08 4,0 21,7 0,13 12,3
4.1 21,6 0,03 0,0 21,7 0,05 2,0 21,7 0,06 3,3
4.2 22,0 0,05 0,0 22,0 0,07 4,3 21,9 0,08 5,4
4.3 22,3 0,06 1,4 22,3 0,08 4,9 22,2 0,08 6,2
5.1 21,6 0,04 0,9 21,6 0,10 8,2 21,6 0,11 10,7
5.2 21,9 0,06 2,8 21,9 0,09 5,2 21,8 0,11 9,6
5.3 22,1 0,05 1,1 21,9 0,08 5,8 21,8 0,10 8,4
6.1 21,7 0,04 0,4 21,6 0,15 13,2 21,4 0,20 20,5
6.2 22,0 0,07 4,2 21,8 0,09 6,9 21,7 0,15 15,1
6.3 22,0 0,08 5,8 21,9 0,08 5,8 21,8 0,14 11,6
7.1 21,7 0,02 0,0 21,7 0,04 0,0 21,8 0,05 1,9
7.2 22,1 0,05 0,9 22,0 0,07 4,3 22,0 0,10 7,9
7.3 22,2 0,05 0,6 22,1 0,08 4,8 22,0 0,10 8,0
8.1 21,6 0,04 1,0 21,7 0,08 5,0 21,6 0,09 8,6
8.2 22,0 0,05 0,9 21,9 0,08 5,4 21,8 0,09 7,1
8.3 22,2 0,05 0,5 22,0 0,09 6,8 21,9 0,09 7,2
9.1 21,8 0,04 0,0 21,7 0,10 8,8 21,4 0,15 16,0
9.2 22,1 0,06 2,7 21,9 0,09 6,6 21,6 0,13 13,8
9.3 22,2 0,07 3,7 22,0 0,08 5,6 21,7 0,13 13,5
117
Table B.2
Measured velocity and temperature values and calculated average draught rating DR for 3
different airflow cases and WITH SUSPENDED CEILING mounting case. The values in
the tables are average values of three replicates made in the same conditions. Point x.1 is
0,1m above floor, point x.2 is 0,6m above floor and point x.3 is 1,1m above floor. The
uncertainties given for the average operative temperature top, average supply air
temperature tsupply and for the average airflow rate V& are expanded uncertainties
(coverage factor k = 2).
Case 4 Case 5 Case 6
top= 22,5 ± 0,2 °C top= 22,2 ± 0,2°C top= 22,4 ± 0,2°C
tsupply=15,1 ±0,3°C tsupply=15,0± 0,1°C tsupply=15,1 ± 0,3°C
V& = 9,7 ± 1,1 l/s V& = 27,6 ± 1,7 l/s V& = 55,9 ± 2,7 l/s
Point Temp v DR Temp v DR Temp v DR

°C m/s % °C m/s % °C m/s %
1.1 21,4 0,02 0,0 21,3 0,05 1,5 21,1 0,07 4,4
1.2 22,0 0,05 1,2 21,8 0,10 7,2 21,7 0,11 9,3
1.3 22,3 0,05 0,5 22,1 0,10 7,3 22,0 0,13 13,0
2.1 21,6 0,03 0,0 21,5 0,12 11,2 21,3 0,09 7,8
2.2 22,2 0,04 0,0 21,9 0,06 2,4 21,7 0,11 10,3
2.3 22,3 0,04 0,0 21,9 0,05 1,1 21,6 0,11 10,1
3.1 21,3 0,03 0,0 21,5 0,09 7,8 21,3 0,12 12,0
3.2 21,9 0,04 0,0 21,9 0,08 5,5 21,6 0,13 13,1
3.3 22,2 0,04 0,0 22,0 0,07 3,7 21,5 0,17 17,1
4.1 21,3 0,02 0,0 21,3 0,03 0,0 21,0 0,07 4,6
4.2 21,8 0,04 0,0 21,8 0,05 1,5 21,6 0,10 8,2
4.3 22,3 0,04 0,0 22,2 0,06 1,8 21,9 0,10 8,7
5.1 21,6 0,02 0,0 21,5 0,10 9,4 21,1 0,13 12,5
5.2 22,2 0,04 0,0 21,9 0,06 2,3 21,5 0,13 11,5
5.3 22,4 0,04 0,0 21,8 0,06 3,1 21,4 0,14 12,3
6.1 21,4 0,02 0,0 21,3 0,06 4,2 21,0 0,23 24,0
6.2 22,0 0,04 0,0 21,8 0,08 5,5 21,4 0,12 11,1
6.3 22,3 0,04 0,0 21,9 0,08 5,4 21,5 0,13 12,9
7.1 21,4 0,02 0,0 21,2 0,02 0,0 21,2 0,05 1,8
7.2 21,9 0,04 0,0 21,7 0,05 1,2 21,8 0,09 7,3
7.3 22,2 0,04 0,0 21,9 0,04 0,0 22,0 0,13 12,1
8.1 21,5 0,02 0,0 21,2 0,03 0,0 21,2 0,09 8,0
8.2 22,1 0,03 0,0 21,7 0,06 1,5 21,6 0,13 12,6
8.3 22,2 0,04 0,0 21,9 0,05 1,2 21,6 0,13 10,9
9.1 21,4 0,02 0,0 21,2 0,03 0,0 21,1 0,15 16,3
9.2 22,1 0,04 0,0 21,9 0,05 1,2 21,4 0,18 19,2
9.3 22,3 0,04 0,0 22,0 0,08 4,7 21,4 0,17 17,4
118
2.8
10 l/s 1,1m level
0,05 0,05 0,05
2.1
0,06 0,06
1,1 m
Room width, m
0,07
0,05
0,06 1.4
0,07 0,08
1,1 m measurement
level
0.7
0,05 0,07
0,05 0,06
0.0
3.9
0.0
0.98
1.95
0.98
Room length, m 2.8
10 l/s 0,6m level
0,05 0,04 0,05 0,06

2.1
0,05
Room width, m
0,6m
0,06
0,05 0,07 1.4
0,06
0,6 m measurement
level 0.7
0,05 0,06
0,05
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8

10 l/s 0,1m level
0,03 0,04 0,05 0,04
2.1
Room width, m
0,1m
0,04
0,03 0,04 1.4
0,1 m measurement
level 0.7
0,02 0,04
0,03 0,04
Figure B.1. Iso- 0.0

velocity profiles in
3.9
0.0
0.98
1.95
0.98
measurement case 1.
Crossed dots on a Room length, m
measurement area mark the measured points. The supply air temperature was +15˚C.
119
2.8
25 l/s 1,1m level
0,09 0,08 0,08

2.1
0,085
Room width, m
1,1 m
0,08
0,08 0,08 1.4
0,085
0.7
1,1 m measurement 0,08 0,08
level 0,09
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
25 l/s 0,6m level
0,08 0,09 0,08 0,07
2.1
0,08
Room width, m
0,6m
0,07 1.4
0,08 0,09
0,6 m measurement
0.7
level
0,07 0,08 0,09
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
25 l/s 0,1m level
0,05 0,06 0,08 0,10 0,11 0,10 0,09

0,10
2.1
0,12
Room width, m
0,08 0,14
0,1m 0,14
0,10
0,05 0,15 1.4
0,12
0,14
0,06
0,1 m measurement 0,12

level 0.7
0,06 0,08 0,10

0,04

3.9
0.98
1.95
0.98
0.0
measurement case 2.
Room length, m
Crossed dots on a
120
2.8
50 l/s 1,1m level
0,13 0,13 0,13

0,12
0,12 2.1
0,13
1,1 m
0,10
Room width, m
0,10 0,11
0,09 0,10 0,12

0,08 0,14 1.4
0,11
level 0,10 0.7
0,10 0,09 0,11 0,12 0,13
0.0
3.9
0.98
1.95
0.98
0.0
Room 2.8
50 l/slength, m
0,6m level
0,14 0,13 0,12
0,13 0,12 2.1

0,12 0,13
0,11
Room width, m
0,6m 0,10 0,12 0,14
0,09 0,10 0,11 0,14 0,15
0,08 1.4
0,13
0,09 0,14
0.7
level
0,10
0,09 0,10 0,12 0,13
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
50 l/s 0,1m level
0,06 0,08 0,09 0,10 0,12 0,13

0,14
2.1
0,16
Room width, m
0,1m 0,18
0,08 0,10 0,12 0,18
0,06 0,14 0,16 1.4
0,20
0,18
level 0,16 0.7
0,05 0,06 0,08 0,10 0,12 0,14 0,15

3.9
0.98
1.95
0.98
0.0
measurement case
Room length, m
3. Crossed dots on
a measurement area mark the measured points The supply air temperature was +15˚C.
121
2.8
10 l/s 1,1m level
0,05 0,04 0,04

2.1
1,1 m
0,045
Room width, m
0,04 0,04
0,04 1.4
1,1m measurement 0.7

level 0,04 0,04 0,04
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
10 l/s 0,6m level
0,05 0,04 0,04
2.1
0,045
Room width, m
0,6m
0,04 0,04
0,04 1.4
0,035
0.7
0,6m measurement level 0,04 0,04
0,03
0.0
3.9
0.98
1.95
0.98
0.0
2.8
10 l/s 0,1m level
0,02 0,025 0,03 0,03
2.1
0,025
Room width, m
0,1m
0,02 0,02
0,02 1.4
0,1m measurement level

0.7
0,02 0,02
0,02
0.0
Figure B.4: Iso-velocity
3.9
0.98
1.95
0.98
0.0
profiles in measurement
Room length, m
case 4. Crossed dots on a
122
2.8
25 l/s 1,1m level
0,10 0,08 0,06 0,05 0,07
2.1
1,1 m 0,09
0,08 0,06
0,08
Room width, m
0,07
0,06 0,07
0,06 0,08 1.4
1,1m measurement
level
0,05
0.7
0,04 0,05 0,06 0,07 0,08
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
25 l/s 0,6m level
0,10 0,08 0,06 0,07 0,08

0,09 2.1
0,08
0,08
Room width, m
0,6m 0,07
0,06
0,06
0,05 0,08 1.4
0,07
0,07
0,6m measurement
0.7
level
0,05 0,06 0,05
0.0
3.9
0.98
1.95
0.98
0.0
2.8
25 l/s 0,1m level-FS
0,06 0,08 0,10 0,12 0,10 0,09
2.1
0,10 0,08
0,04 0,06
Room width, m
0,1m
0,03 0,10 0,06 1.4
0,08
0,1m measurement 0,06

0,04 0,04
level 0,04 0.7
0,02 0,03 0,03

3.9
0.98
1.95
0.98
0.0
measurement case 5.
Crossed dots on a Room length, m
123
2.8
50 l/s 1,1m level
0,13 0,12 0,11 0,13 0,14 0,17
0,12 0,16 2.1

0,12
1,1 m 0,12 0,15
Room width, m
0,15
0,13 0,14
0,10 0,13 1.4
0,11 0,15
1,1m measurement 0,16 0,15

0,12 0,12
level 0.7
0,13 0,13 0,14 0,16 0,17
0.0
3.9
0.98
1.95
0.98
0.0
Room length, m 2.8
0,11 0,12 0,13
2.1
Room width, m
0,6m
0,11 0,13 0,12

0,10 1.4
0,12 0,14 0,13
0,15
0,16 0,15
0,6m measurement 0,16
0,17
level 0,17 0.7
0,09 0,10 0,11 0,13 0,14 0,16 0,18
0.0
3.9
0.98
1.95
0.98
0.0
2.8
0,07 0,08 0,09 0,10 0,12

0,14 2.1
0,08
0,14 0,16
0,18
Room width, m
0,1m 0,10 0,20

0,13 0,22
0,07 0,18
0,22 0,23 1.4
0,12
0,16 0,20 0,20
0,1m measurement 0,06 0,08
0,14 0,18
level 0.7
0,16
0,05 0,06 0,08 0,09 0,10 0,12 0,14 0,15
0.0
Figure B.6. Iso-
3.9
0.98
1.95
0.98
0.0
measurement case 6. Room length, m
Crossed dots on a
124
C Questionnaire
Questionnaire nr:_________________
Data:__________________
Indoor climate evaluation – EDIT- building at Chalmers

The aim of this questionnaire is to evaluate the work of the renovated air conditioning
system in EDIT-building at Chalmers. CIT Energy Management AB with co-operation
with Building Services Engineering- dep. of Energy & Environment at Chalmers is
carrying out a questionnaire which is commissioned by Akademiska Hus AB. The
purpose of this questionnaire is to study the building users’ perceptions towards
their indoor environment in general. This questionnaire was first carried out in spring
2004 and is repeated now after the summer 2005.
The air conditioning system in EDIT-building at Chalmers was fully renovated in

2003 and taken in operation in September 2003. After the renovation of the air-
conditioning system, new supply air diffusers were installed to the ceiling of each
office room. During the winter time 2003/2004 some small problems with the system
work appeared, which fortunately have been solved by now. After the summer 2004,
the system has worked in a way it was planned.
A place for personal comments is given under the question 11 at the last page.
This questionnaire refers to the indoor environment during the previous summer
period. Think about the situation you have had in your room when the outdoor
conditions were warm, which has been the period from May till September.
The questionnaire is given out in the morning and collected during the end of the
day or before the afternoon next day
We would like to stress that answering to this questionnaire is voluntary and the
information gather will be handled confidentially.
Sincerely yours,
Lennart Jagemar Mari-Liis Maripuu

Associate professor PhD Student
125
Air temperature
1a. How do you feel about the room temperature during this period of the year?
Cold cool Slightly Neutral Slightly Warm Hot

cool warm
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
Extremely Very uncomfortable Slightly Comfortable

❑1 ❑2 ❑3 ❑4 ❑5
1c. Please mark, how would you like to have the temperature during this period of the
year?
Much Colder Slightly It is good Slightly warmer much

colder colder as it is warmer warmer
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
Air movement (Draught)
2a. How do you perceive the air movement (draught) during this period of the year?
Very Low Slightly Neither Slightly High Very high

low low high nor High
low
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
2c. Please mark, how would you like to have the air movement this period of the
year?
Much Lower Slightly It is good Slightly Higher Much
Lower Lower As it is Higher higher
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
Air humidity
3a. How do you perceive the air humidity during this period of the year?
low
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
126

❑1 ❑2 ❑3 ❑4 ❑5
3c. Please mark, how would you like to have the air humidity during this period of the
year?
Much Dryer Slightly It is good Slightly More Much more

dryer dryer As it is more humid humid
humid
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
Noise level
4a. How do you perceive the noise level in your room during this period of the year?

low
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
4c. Please mark, how would you like to have the noise level during this period of the
year?
Much Lower Slightly It is good Slightly Higher Much

Lower Lower As it is Higher higher
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
4d. If you feel that the noise level is in some extent uncomfortable- what does it
depend on?
Ventilation ❑1
Lightning ❑2
Office equipment i.e. ❑3
computer
Conversation, talking ❑4
Noise outside ❑5
❑6
Other Other
sources:__________________________________
127
Lights- Office lightning
5a. How do you feel about the room lightning during this period of the year?
Very Weak Slightly Neither weak Slightly Strong Very

weak Weak nor strong strong strong
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
5c. Please mark, how would you like to have the office lightning during this period of
the year?
Much Weaker Slightly It is good Slightly Stronger Much

Weaker Weaker As it is Stronger stronger
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
5d. If you feel that the office lightning is in some extent uncomfortable- what does it
depend on?
Dazzling 1
Unequal lightning 2
Other 3 Name other:___________________________________________
Lights – Day light
6a. How do you feel about the day light at your workplace during this period of the
year?
Very Weak Slightly Neither Slightly Strong Very

weak Weak weak nor strong strong
strong
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
128
6c. Please mark, how would you like to have the day light during this period of the
year?
Much Weaker Slightly It is good Slightly Stronger Much

Weaker Weaker As it is Stronger stronger
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7
6d. If you feel that the day light is in some extent uncomfortable- what does it depend on?
Dazzling 1
Reflection on the 2
computer screen
Shadow effects 3
4
Other Name other: ______________________________________
Indoor air quality
7a. How do you feel about the indoor air quality (air quality refers to
dust/odours/stuffy) in the air) in the room during this period of the year?
Very Bad Slightly Neither bad Slightly Good Very

bad bad nor good good good
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
7c. If you feel that the air quality is in some extent uncomfortable- what does it
depend on?
Dust 1
2
Odours
3
Stuffy
4
Other Name other:_________________________________________
7d. If you feel that the air quality is in some extent uncomfortable – does it appear
during some particular time?
In the mornings At lunchtime In the afternoon In the evening Saturday/Sunday

- holidays
❑1 ❑2 ❑3 ❑4 ❑5
129
Self influence to the indoor climate
8a. In what extent do you think you could influence following things?
Certain
Not at all A little A lot entirely
amount
Temperature ❑1 ❑2 ❑3 ❑4 ❑5
Noise ❑1 ❑2 ❑3 ❑4 ❑5
Lightning ❑1 ❑2 ❑3 ❑4 ❑5
Day light ❑1 ❑2 ❑3 ❑4 ❑5
Indoor air quality ❑1 ❑2 ❑3 ❑4 ❑5
8b. Which needs do you think you have to change from the following?
Not at all A little moderate Big Very big

Temperature ❑1 ❑2 ❑3 ❑4 ❑5
Noise ❑1 ❑2 ❑3 ❑4 ❑5
Lightning ❑1 ❑2 ❑3 ❑4 ❑5
Day light ❑1 ❑2 ❑3 ❑4 ❑5
Indoor air quality ❑1 ❑2 ❑3 ❑4 ❑5
Indoor environment
9a. How do you evaluate your indoor environment in generally during this period of
the year?
Very Bad Slightly Neither bad Slightly Good Very good

bad Bad Nor good Good
❑1 ❑2 ❑3 ❑4 ❑5 ❑6 ❑7

❑1 ❑2 ❑3 ❑4 ❑5
Working environment [According to WHO: All factors – biological, medical,

physiological, social and technical- which in the work situations an in the working place
and its surroundings influence the individual]
10a. In what extent do you feel your work assignments to be interesting and
stimulating?
Not at all A Little Certain amount Quite a lot Entirely
❑1 ❑2 ❑3 ❑4 ❑5
130
10b. In what extent can you influence your working conditions?
Not at all A Little Certain amount Quite a lot Entirely
❑1 ❑2 ❑3 ❑4 ❑5
Other
11. Other comments regarding indoor environment in your work space
_________________________________________________________________
_________________________________________________________________
_________________________________________________________________
Background information
12. Are you a man or a woman? Man Woman

❑1 ❑2
13. Which year are you born
in?____________
14. How long have you worked in your present working space? ___________months
15. How many work places, including your own, are there in your office room?
❑1 ❑2 ❑3 ❑4 ❑5
Have you answered to this type of questionnaire before (carried out Yes 1
in winter) No 2
131

MARI LIIS MARIPUU Licentiate Thesis

Uploaded by

Copyright:

Available Formats

MARI LIIS MARIPUU Licentiate Thesis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MARI LIIS MARIPUU Licentiate Thesis

Uploaded by

Copyright:

Available Formats

THESIS FOR THE DEGREE OF LICENTIATE OF ENGINEERING

Adapting Variable Air Volume (VAV) Systems for Office

Building Services Engineering

Technical report D2006:02

One example of uncomplicated system configuration, which in theory has properties

Gothenburg, May 2006

3 FIELD MEASUREMENTS – TWO CASE STUDIES…………………...27

4 LABORATORY MEASUREMENTS …………………………..…………37

5 RESULTS AND DISSCUSSION….. …………………………..…………43

6 EVALUATION OF THE HEAT GAINS IN A VAV DUCT SYSTEM ….79

7 CONCLUSIONS AND FUTURE RESEARCH ………………………….99

dA surface area of a duct element, [m2]

ACH air changes per hour

300 Electricity Heating

Heat generation- Heat losses, W/m

CAV cooling capacity

VAV cooling capacity

Heat loads Heat loads

1.1.3 The supply air temperature

1.2 Objectives and scope

The methodology comprises three different ways of studies:

1.4 Outline of the thesis

Appendix B presents some of the results from the laboratory measurements as a

1.5 Literature review

2.1.2 Indoor climate problems associated with VAV

Insufficient supply of outdoor air

According to previous studies, this complaint may occur in circumstances where

The draught rate can be calculated by the equation:

DR = (34 − t a )(va − 0,05) 0,62 (0,37 ⋅ va ⋅ Tu + 3,14) [%] (2.1)

2.2 Technical aspects

V&min V&design Airflow rate,V&

V&min V&design Airflow rate,V&

A pressure dependent VAV diffuser/unit is a device where the airflow control is

Pressure independent air

• Demand based airflow control by room temperature, by room air quality, by

• It should be possible to supply air with a low temperature, +15°C or lower,

• The cost must be acceptable from a life cycle point of view.

2.4 Demands for other VAV system components

2.5 Energy aspects

2.5.1 Electrical energy savings

∆ ptot = f (const + V&VAV

AHU ∆pSP =∆pduct

V&design Airflow rate, V&

η tot = f (V& , ∆ptot ) (2.6)

Fraction of design airflow rate

Total drive system efficiency

A schematic picture of the energy consumption of a fan in a relation with varying

Airflow rate, V& V& design

Electrical energy efficiency of a ventilation system is commonly characterized by the

The variation in SFP value can be written as a function of airflow as follows:

Airflow rate, V& V& design

2.5.2 Thermal energy savings

Q& h = V&VAV ⋅ ρ ⋅ c p ⋅ (t sup ply − t rec ) [W] (2.10)

Q& c = V&VAV ⋅ ρ ⋅ ∆h [W] (2.11)

3.1 Case study 1- EDIT-building at Chalmers

3.2 Case study 2- Academicum and

3.3.2 Energy consumption measurements

During the energy consumption measurements following parameters were monitored:

3.4 Indoor climate evaluation - questionnaire

An example of the questions asked for evaluating one environmental parameter is