Energy and Buildings 55 (2012) 765–778

Contents lists available at SciVerse ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

A methodology for economic efficient design of Net Zero Energy Buildings

M. Kapsalaki a,∗ , V. Leal a , M. Santamouris b
a
IDMEC, Institute of Mechanical Engineering, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
b
National and Kapodistrian University of Athens, Greece, Physics Department Section of Applied Physics, Physics Department, Build. Physics 5, 15784 Athens, Greece

a r t i c l e i n f o a b s t r a c t

Article history: This work developed a methodology and an associated calculation platform in order to identify the eco-
Received 14 August 2012 nomic efficient design solutions for residential Net Zero Energy Building (NZEB) design considering the
Received in revised form influence of the local climate, the endogenous energy resources and the local economic conditions. One
28 September 2012
case study of a detached house for 3 climates was analyzed with the tool developed in order to gain
Accepted 10 October 2012
insights on the economic space of NZEB solutions and the influence of the climatic context. A methodol-
ogy for assisting the choice of economically efficient NZEB solutions from the early design stage is now
Keywords:
available. Its use in practice may be of great relevance as the results showed that the differences between
Net Zero Energy Building
Sustainable building
an economically efficient and economically inefficient NZEB can be over three times both in terms of
Energy efficient building initial and life cycle cost.
Life cycle cost © 2012 Elsevier B.V. All rights reserved.
Microgeneration

1. Introduction
Buildings are seen as a key-part of the needed transition toward
sustainability in its energy dimension. This derives from the fact
that the buildings sector represents between 30% and 40% of the
demand of final energy in most developed countries. Given that
in buildings there is a high use of electricity, these figures often
convert to about or more than 40% in primary energy and energy-
related CO2 emissions creating problems in terms of greenhouse
effects and climate change, but also for many countries of security
of supply and/or of economic competiveness and environmental
compatibility.
A radical approach for the mitigation of the energy demand as
regards the buildings sector is the concept of the NZEB [1–6]. This
concept started to appear in the literature as an evolution of very
energy-efficient buildings, and it requires that the building that has
zero energy balance on an annual basis. Even if it may absorb energy
from the electric or gas grids at some times, it has local production
Abbreviations: Al, aluminum; AFUE, Annual Fuel Utilization Efficiency; BP, price
of electricity bought from the grid; CDD, Cooling Degree Days; COP, Coefficient of
Performance; CS, central scenario; DHW, domestic hot water; EF, energy factor;
HDD, Heating Degree Days; IC, initial cost; LCC, life cycle cost; LIC, lowest initial cost;
LLCC, lowest life cycle cost; NZE, Net Zero Energy; NZEB, Net Zero Energy Building;
NZEBs, Net Zero Energy Buildings; PB, payback; PV, photovoltaic; RCHP, reversible
cycle heat pump; RCHP class A, reversible cycle heat pump class A; SC, solar collector;
SP, price of electricity sold to the grid; TB, thermal break; TPO, total power output.
∗ Corresponding author.
E-mail addresses: pds07005@fe.up.pt, kapsalakimaria@gmail.com
(M. Kapsalaki).
systems that export energy into the electric grid at other times so
that the annual balance is about null.
Net Zero Energy Buildings (NZEBs) are becoming an important
policy option in many countries and regions of the world. In the
European Union (EU), the Energy Performance of Buildings Direc-
tive (EPBD) approved in 2002 [7] called for the assessment and
publicity of the energy performance of all new buildings as well as
of those undergoing major refurbishments or entering the market
for rental or selling. This directive was later updated by a recast
approved in May 2010 [8]. This new version requires that as of
31 December 2020, new buildings in the EU will have to require
‘nearly zero’ energy and the energy will be ‘to a very large extent’
from renewable sources.
There are now some examples of NZEBs already built, which
show that they are achievable [9–13]. However, it has not been
proven that the design choices are, besides technically effective,
also efficient from the point of view of the use of economic
resources. For instance, it is important to develop methodologies
that allow building designers to identify the combinations of design
variables (from potentially millions of possible combinations) that,
while ensuring the achievement of the energy and environmental
targets established, also have near-optimal lowest LCC, or lowest
IC, or a good compromise between IC and LCC. That is the sub-
ject of this work, which explains the process of development of a
methodology and a connected tool for this purpose and exempli-
fies its potential use. It expands the scope of previous works on
the identification of cost-optimal solutions for nearly zero energy
buildings [14], of multi-objective optimization for low-emission
cost-effective dwellings [15] and of minimizing the life-cycle cost
of non-NZEB detached houses [16].

0378-7788/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.enbuild.2012.10.022
766 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
Fig. 1. Breakdown of EU-27 residential electricity consumption, year 2007 [13].
This tool can be used to decide the most adequate design solu-
tions that lead to a null energy annual balance (made in terms of
final energy), for any residential building in any part of the world,
provided that the adequate local climatic and economical inputs
are inputted. For the sake of clarity, and because many variant def-
initions now exist, the baseline concept of NZEB for this study is
that which is bidirectionally connected to the electric grid and has
the microgeneration equipment integrated in the building itself.
The paper is laid out as follows: Section 2 identifies and
discretizes the key design variables; Section 3 addresses the cal-
culation methodologies that were used in order to characterize the
energy demand and supply of the building as well as the economic
indicators of each design alternative and provides a description of
the computer program developed; Section 4 provides a first appli-
cation example of the methodology, while conclusions are drawn
in Section 5.
2. Analysis model and software
2.1. Energy uses in residential buildings
Characterizing the energy demand of a residential building
involves firstly identifying which are the specific energy end uses
involved. According to statistics, the energy end uses with rele-
vance in the residential sector are: space heating and cooling, water
heating (domestic hot water), lighting, cooking, refrigeration and
other appliances (including computers and multimedia devices).
The pie charts of Figs. 1 and 2 show the estimated energy use break-
down of the residential sector in Europe in 2007 [17] and the United
States in 2005 [18] respectively.
Once the end uses are identified, the electricity and gas con-
sumption profiles can be assessed and the overall domestic load
can be characterized. In this way the needs to import energy to
the building can be estimated and in the case of NZEBs the needed
Fig. 2. Breakdown of residential sector energy use in United States [12].
equivalent amount of energy (from local microgeneration to export
to offset the imports) can be determined.
2.2. Building design variables and design alternatives
Each major end use that shapes the energy demand can be influ-
enced by a number of design variables, and typically each building
design variable has a wide range of possible values or choices. Fig. 3
shows the dependence of each end use on each design variable.
Each combination of design variables leads to a certain total
yearly energy demand, under standard building use conditions,
which requires a certain amount of building-integrated renewables
to offset the demand on a yearly balance basis. It therefore is of
interest to characterize the full spectrum of possible combinations
of variables, in order to identify those that have expected lower ini-
tial costs and those that have lower life-cycle costs, but also how
distant from ‘the best’ some other solution (e.g. one preferred due
to architectural/non-energy criteria) may be. The expression ‘build-
ing design alternative’ will be adopted here as referring to a certain
combination of values or options for the design variables of Fig. 3.
Therefore, if two combinations differ in at least one value/option of
a variable, then they are considered as different design alternatives.
2.3. Discretization of the design variables
Many of the building design variables can vary in a continuous
manner (e.g. the wall thermal insulation U-value), and others can
take finite but often very high number of options (e.g. efficiency
of the heating equipment, which has thousands of models in the
market). In order to treat the problem it is however necessary to
bring to a finite and reasonable dimension. This was achieved by
discretizing the space of variation of each design variable and, as
consequence, the space of building design alternatives. The dis-
cretization of each variable is described in the next sections, and
was made with the criteria of balancing a good coverage of the
space of variation with keeping the number of options low, since
in combinatory calculus the total number of alternatives grows
exponentially with the number of options for the design variables.
2.3.1. Thermal insulation level
Thermal insulation was characterized by four different levels.
These levels are represented by the U-value, which ranges from
0.2 W/m2 K (AECB requirement [19]) up to 3 W/m2 K (represents
an uninsulated wall), with two intermediate levels of 0.8 W/m2 K
(roughly the maximum allowed by the current Portuguese ther-
mal building regulation (RCCTE) [20]) and 0.35 W/m2 K (maximum
for compliance with the part L of the 2006 Edition of the Building
Regulations for England and Wales [21]). Besides the U-value itself,
each level is characterized also by a corrected U-value which is the
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778 767

Fig. 3. Energy uses versus design variables.

sum of the U-value plus a default surcharge U taking into account
the effect of thermal bridges according to the ISO 13790:2008
[22]. This results in corrected U-value levels of 0.30, 0.45, 0.80 and
3.0 W/m2 K.
2.3.2. Leakage level and ventilation type
The infiltration levels chosen to characterize the airtightness of
the residence, expressed in air changes per hour (ach−1 ) on average,
were 0.1, 0.6 and 1.1 ach−1 . Three types of ventilation were consid-
ered as possible alternatives; natural, mechanical and mechanical
with heat recovery. The ventilation rate per occupant considered in
the case of mechanical ventilation was 35 m3 /h and the heat recov-
ery efficiency 75%. Table 1 shows all the combinations of infiltration
and ventilation systems together with the values of air changes
for each one. In the case of heat recovery, it is only applied to the
air ventilated through the mechanical system, not to the air from
infiltration.
2.3.3. Window and shading types
A range of different windows was chosen to reflect the vari-
ety found in the market, considering different climate zones. The
options considered single, double and triple glazing, framed with
or without thermal breaks, and glazing with or without low-e films,
according to the combinations shown in Table 2.
The range of shading options was clustered into three types:
none, indoors or outdoors. The g-values for the cooling season were
calculated as sum of the 30% of the g-value of the glazing plus
70% of the g-value of the window considering the blinds closed
[20]. During the heating season the blinds were considered open
(Table 2).
2.3.4. Glazed area
The glazed area design variable was represented as a ratio to the
heated useful area, and it was discretized into three levels: 10%, 20%
or 40%.

Table 1
Air changes per hour default values.

Infiltration (ach−1 ) Ventilation

Natural Mechanical Mechanical with heat recovery

Low (0.1a ) 0.1 + (35 × occupants)/V 0.1 + 0.25 × (35 × occupants)/V

Medium 0.6 0.6 + (35 × occupants)/V 0.6 + 0.25 × (35 × occupants)/V
High 1.1 1.1 + (35 × occupants)/V 1.1 + 0.25 × (35 × occupants)/V
a
The combination of natural ventilation with low infiltration is not advised and therefore not considered as a possibility in this study.
768 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778

Table 2
U-Values and g-values of each window type and shading [5,20,22,43].

Window type Shading

Inside Outside None

2◦ 2◦
U (W/m C) gh gc U (W/m C) gh gc U (W/m2 ◦ C) gh gc

Single glazed clear aluminum (Al) 5.68 0.70 0.65 5.68 0.70 0.30 5.68 0.70 0.70
Double glazed clear Al 4.83 0.63 0.62 4.83 0.63 0.27 4.83 0.63 0.63
Double glazed clear Al with thermal break (Al + TB) 3.58 0.63 0.62 3.58 0.63 0.27 3.58 0.63 0.63
Double glazed low-e, AL + TB 2.73 0.63 0.60 2.73 0.63 0.25 2.73 0.63 0.63
Triple glazed clear Al 2.85 0.50 0.61 2.85 0.50 0.26 2.85 0.50 0.50
Triple glazed clear, Al + TB 2.49 0.50 0.61 2.49 0.50 0.26 2.49 0.50 0.50
Triple glazed low-e 2.23 0.50 0.55 2.23 0.50 0.20 2.23 0.50 0.50
Triple glazed low-e, Al + TB 1.86 0.50 0.55 1.86 0.50 0.20 1.86 0.50 0.50

2.3.5. Orientation Table 4

Energy factors – DHW systems [24,44].
Two options were considered for the facade orientation: (i) one,
which gives preference to the South orientation for the installation EF Lifetime (years)
of the windows/glazings, representative of the passive solar design Heat pump 2.1 10
[23]. In this case it is assumed that 70% of the glazed area is fac- Conventional storage electricity 0.95 13
ing south while north, east and west glazed areas have each a 10% Conventional storage gas 0.67 13
share over the total; (ii) another one where there is no preferential Conventional storage condensing 0.94 13
Tank less or instantaneous gas 0.82 20
orientation, and the glazings are uniformly distributed through all
Solar collector 0.12 19
orientations.

2.3.6. Thermal mass
Thermal mass/inertia was discretized into three levels. These
were: low (useful thermal mass smaller than 150 kg/m2 of floor
area), medium (useful thermal mass between 150 and 400/m2 of
floor area) and high (useful thermal mass bigger than 400/m2 of
floor area) [20].
The quantitative parameters that characterize each system
include the energy factor (EF) of the DHW system (ratio of use-
ful energy output from the water heater to the total amount of
energy delivered to the water heater), the expected lifetime and
the annual energy produced by the solar collector. Table 6 shows
the EF considered for each DHW system (Table 4).

2.3.7. Heating and cooling system 2.3.9. Lighting and appliances

The heating system options considered consist of gas boiler, con-
densing gas boiler, electric radiator, conventional reversible cycle
heat pump (RCHP) and class A RCHP. For the cooling systems the
options considered were conventional RCHP and class A RCHP. The
‘class A RCHP’ can be seen as representing ground-source heat
pumps too (with the COP referred to the outdoor air temperature).
Each option is characterized by the Annual Fuel Utilization Effi-
ciency (AFUE), which is the ratio of heat output to the total energy
consumed by the boiler [24], or the Coefficient of Performance
(COP) for the heat-pump based systems, and the typical lifetime
of each system. For the case of the RCHP and the RCHP class A, two
values for the COP were considered for the heating and the cooling
season. Table 3 shows the values considered for each one of the
systems.
2.3.8. Domestic hot water system (DHW)
The options considered to provide domestic hot water for the
building are a heat pump water heater, a conventional storage elec-
tric boiler water heater, a conventional storage gas boiler water
heater, a conventional storage gas condensing boiler water heater
and a tankless or instantaneous gas water heater. These hot water
systems can also serve as auxiliary systems with the co-existence
of a solar collector (SC).
Two alternatives were considered for lighting and appliances:
high efficiency or average efficiency. The electric appliances consid-
ered include a refrigerator & freezer, clothes dryer, a dishwasher,
an oven/cooker, a TV and a desktop PC. Their reference annual con-
sumptions, for each of the efficiency alternatives are presented in
Table 5 [25].
2.3.10. Microgeneration system
The range of available microgeneration technologies considered
the most widely used for residential on site electricity generation:
PV panels and small wind turbines. The PV modules were charac-
terized by their efficiency, in two levels: 12% average efficiency and
18% high efficiency.
Regarding the wind turbines, three types were considered:
400 W, 1.8 kW and 2.4 kW. This was due to the fact that the power
curves are considerably different according to the rated power.
The annual energy production for each one of the systems
is computed based on the climate data of each location (solar
radiation and wind speed) as well as the ‘site description’, which
Table 5
Consumptions per appliance (kWh/year) [25].
Appliances and lighting Average annual energy consumption (KWh/year)
(per dwelling)

Table 3 Average efficiency High efficiency

AFUE or COP and lifetime of the heating systems considered [24,44]. Refrigerator & freezer 451 219
Washing machine 184 136
AFUE or COP Lifetime (year)
Clothes dryer 347 174
Boiler 0.85 25 Dishwasher 234 163
Condensing boiler 0.94 25 Oven/cooker 301 226
Electric radiator 0.99 8 TV 400 337
RCHP 3.4 (summer) 2 (winter) 16 Desktop PC 276 100
RCHP class A 5.5 (summer) 2.6 (winter) 16 Lighting 487 184
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
is also user defined and includes the following options: rural, rural
with obstacles and urban. This serves for the correction of wind
speed for different heights and for different ground obstructions
(see Section 3).
2.3.11. Occupants
The number of occupants needs to be defined so as to determine
the hot water needs as well as the thermal internal gains. For the
software to be presented in Section 3.11, a default value of four
occupants is considered.
The occupancy schedule that is assumed is fixed throughout the
year; for children, the building is assumed to be occupied for 38
weeks under normal schedule, plus 14 weeks under the holiday
schedule. Adults have a different occupancy schedule: 48 weeks
under the normal schedule and 4 weeks under the holiday schedule.
2.3.12. Climate
A database with nine climates is pre-built in the design tool to be
presented, but new ones can be added. Each climate is characterized
by the following parameters: the Heating Degree Days (HDD) for a
base temperature of 18 ◦ C [26], the average monthly solar energy
incident on a vertical surface facing the south during the heating
season, the total solar radiation per orientation during the cooling
season, the duration of the heating season [26], the average outdoor
air temperature during the cooling season, the heating and cooling
design outdoor air temperatures and the average hourly wind speed
for every hour of the year.
3. Energy demand and supply characterization
This section intends to choose and present calculation algo-
rithms for computing the energy uses identified in Section 2. As
much as possible it was intended to use algorithms already avail-
able in the literature rather than developing new ones. The choice
of these algorithms obeyed essentially to the criteria of achiev-
ing a good trade-off between accuracy and computation time. Due
to the need to analyze millions of solutions, hourly or sub-hourly
dynamic computation methods were ruled out. For heating and
cooling, methods based on the ISO13790:2008 [22] and on the
RCCTE [20,27] were preferred. The main challenge was the devel-
opment of a quick algorithm for estimating the cooling peak load.
3.1. Useful heating needs calculation
The buildings’ annual useful energy need for space heating is
calculated as given by Eq. (1) [20].
Qenv + Qair − nh ∗ Qg
HNusf = (1)
A year needed to keep the indoors at the prescribed comfort tem-
where HNusf is the annual useful energy need for space heating
(kWh/m2 year), Qenv is the heat loss through the envelope due to
conduction (kWh/year) in the heating season, Qair is the heat loss
resulting from the air renewal (kWh/year) in the heating season,
nh is the gain utilization factor due to the effect of thermal inertia,
Qg is the total heat gain due to lighting, equipment and occupants,
plus the solar gains through the glazed elements (kWh/year) in the
heating season and A is the buildings’ useful-heated area (m2 ).
3.1.1. Heat loss through the envelope
The heat loss through the envelope (Qenv ) is calculated as the
result of the sum of the heat losses through the walls (Qwalls ), roof
(Qroof ) and glazed areas (Qglz ) in contact with the exterior and the
heat loss through the walls in contact with non-heated areas and/or
adjacent buildings (Qadj ) and is expressed in kWh/year. These losses
result from the temperature difference between the internal and
external building environment.
769
The elements in contact with the exterior include walls, roof and
glazed areas. The heat loss resulting from the walls, from the roof,
from the glazed elements as well as the heat loss caused by the tem-
perature difference to adjacent building and/or non-heated areas
were calculated following the methodology of the ISO13790:2008
in the form adopted by the RCCTE [20].
3.1.2. Heat loss resulting from air renewal
Heat loss during the heating season were calculated as previ-
ously following the methodology of the ISO 13790:2008 [22] in the
form adopted by the RCCTE [20] The sum of the rate of mechanical
ventilation and infiltration rates (ach−1 ) were taken into account
as follows by Eq. (2).
 
V̇NV V̇MV
Rph = + ∗ (1 − nHR ) (2)
V V
with V the volume of the building (m3 ), Rph the sum of the rate
of mechanical ventilation and infiltration rates (ach−1 ), V̇NV the air
flow rate due to natural ventilation (m3 /h) and V̇MV the air flow rate
due to mechanical ventilation (m3 /h) and nHR the efficiency of the
heat recovery system (–) in the case that it exists.
3.1.3. Thermal heat gains
In the RCCTE/ISO13790 methodology [20,22], the heat gains of
the building in the heating season include two components: the
heat gain resulting from lighting, equipment & occupants and the
solar gain through the glazed areas. The calculation of the thermal
internal gains from occupants, lighting and equipment during the
heating season as well as the thermal solar gains from the solar
elements were based on the RCCTE/ISO13790 methodology [20,22]
and are described in detail in [28].
3.1.4. Gain utilization factor for heating
When a building has low thermal inertia, it can happen that
solar or internal heat gains concentrated in a short period reach the
indoor air very quickly and cause overheating. In such a case part of
the gains lose their useful effect. This accounting is done through a
gain utilization factor for heating, whose value depends on the class
of thermal inertia of the building [20]. The dimensionless gain uti-
lization factor for heating (nh ) is a function of the heat-balance ratio
and a numerical parameter that depends on the building thermal
inertia [20].
3.2. Peak heating needs
Peak design heating load calculations determine the maximum
rate of mechanically supplied heating energy needed during the
perature (20 ◦ C in the RCCTE). In this work it is needed to estimate
the needed size of the heating equipment and estimate its cost. A
possible way of estimating this load is to perform the heat balance
for the coldest hour of the year, assuming no internal and no solar
heat gains, under steady state conditions [29]. It is known that by
ignoring thermal inertia effects, this method leads to some overesti-
mation of the heating needs. On the other hand, it is also known that
designers usually like to keep some safety margin for this equip-
ment. Because of this and, also because of the good compatibility
of the method with the unavailability of detailed data in the early
design stages, the method was adopted. Space heating load KWPh
was calculated by computing the heat transfer rate through the
building envelope elements plus the heat that is required due to
outside air infiltration or ventilation [29], through Eq. (3):
KWPh = [(U + U) ∗ Awalls + (U + U) ∗ Aroof + Uglz ∗ Aglz
+ 1.23 ∗ 103 ∗ Rph ∗ V/3600] ∗ T1 (3)
770 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
where T1 is the temperature difference between the indoor
design temperature and the dry bulb temperature corresponding
to 99.6% annual cumulative frequency of occurrence in the heating
season (◦ C). The number 1.23 represents the air heat sensible factor
(W/m3 s ◦ C).
3.2.1. Useful cooling needs
The buildings’ annual useful energy need for space cooling
(CNusf ) is calculated as given by Eq. (4), expressed in kWh/m2 year:
1 − nc
CNusf = Qgn,c ∗
a according to Eq. (5).
⎡ ⎤
(U + U) ∗ Awalls + (U + U) ∗ Aroof + Uglz ∗ Aglz + 1.23 ∗ 103 ∗ Rph ∗
KWPc = ⎣
qi ∗ A + Fg ∗ gc ∗ Sr ∗ Aglz
where Qgn,c is the total gross heat gains (kWh/year), nc is the gains
utilization factor (–) and A is the useful area (m2 ).
Despite more recently it has been shown [27,30] that the equa-
tion adopted in the final version of EN 13790 is slightly more precise
than this while in [27] it is also shown that this one correlates very
well to the results obtained though dynamic hourly simulation.
3.3. Heat sources
Heat sources account for internal gains, solar gains through
transparent elements (windows, skylights) and solar gains through
opaque elements (walls, roof).
The internal heat gains caused by occupants, equipment and
lighting, the solar gains through the transparent surfaces-glazed
areas and the solar gains through the opaque surfaces were calcu-
lated according to the RCCTE/ISO13790 methodology [20,22].
3.4. Peak cooling needs calculation methodology
Peak design cooling load calculations determine the maximum
rate of cooling energy transfer needed at any point during the cool-
ing season in order to maintain the indoor temperature below the
cooling set point. In this work it is needed to estimate the needed
size of the cooling equipment and estimate its cost. Typically, peak
cooling load computing is much more complicated than peak heat-
ing load computing, since the inertia effects are more important
than for the peak heating load and usually must be taken into
account. However, since a dynamic method would not be com-
patible with the requirements of an algorithm that must be very
fast and require little input data, a method was developed here,
Fig. 4. Linear regression of the peak cooling load calculation versus simulation.
based on steady state, under the most severe conditions of the year
(hottest day and highest outdoor temperature), in analogy with
the method of the heating season. Due to the fact that the above
method is very likely to overestimate the cooling load, a correction
factor will be introduced, based on a comparison with the results
of dynamic simulation for a numerous set of buildings.
Space cooling load was therefore calculated in a first phase by
computing the heat transfer rate through the building envelope
elements plus the heat that is transferred due to outside air infiltra-
tion, the internal gains (occupants, equipment and lighting) and the
solar heat gains through the glazed areas, in a steady-state method,
(4)
V
+
3600 ⎦ ∗ T2 (5)
where T2 is the temperature difference between the dry bulb tem-
perature, corresponding to 99.6% annual cumulative frequency of
occurrence (cooling season) and the inside design temperature (◦ C),
and Sr is the maximum solar insolation during the cooling season
(W/m2 ).
In a second stage of the method, the cooling load was computed
by both Eq. (9) and by detailed dynamic thermal simulation using
ESP-r [31]. This process included six climate scenarios (Stockholm,
London, Paris, Prague, Rome and Lisbon) and six building design
alternatives. The graph of Fig. 4 shows the result of the conducted
calculations according to Eq. (9) versus the results of the dynamic
simulations with ESP-r (Fig. 4).
As expected, the values obtained by the steady-state method
were significantly overestimating the peak load. The coefficient of
determination of the correlation has a reasonable value of 0.72. The
calculation of the peak cooling load KWPc including this correction,
results in Eq. (6).
KWPc,final = 0.0593 + 0.5159 ∗ KWPc (6)
3.5. Useful energy needs for domestic hot water
Useful energy needs for domestic hot water (DHWusf ) were esti-
mated based on Eq. (7) [20]:
HWN ∗ cpw ∗ TDHW ∗ 365
DHWusf = (7)
3, 600, 000
where HWN are the needs for hot water per day (RCCTE consid-
ers a default value of 40 l per person), cpw is the specific heat of
the water and TDHW is the temperature difference between the
water leaving for consumption and the water entering the system
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
Table 6
Data for the calculation of the annual energy produced by the thermal collector [32].
Data needed
Solar system efficiency nSC
Zero loss efficiency n0
1st order heat loss coefficient
2nd order heat loss coefficient
Collectors mean fluid temperature Tm
to be heated. For the purpose of this work, three temperature differ-
ences were assumed, depending on the climate (for the climate of
Stockholm the temperature difference considered is 55 ◦ C, for the
climates of Brussels, Wien and Venice 50 ◦ C and for the climates of
Porto, Lisbon, Funchal, Iraklion and Athens 45 ◦ C).
3.6. Total needs of final energy
The needs of final energy for heating, cooling and DHW were
estimated by dividing the useful energy needs by the efficiency
of the systems providing each of the services, according to Eqs.
(8)–(10) respectively:
HNusf
HNfinal =
nHS
CNusf
CNfinal =
nCS
DHWusf
DHWfinal =
nDHWS
where nHS is the efficiency of the heating system, nCS is the effi-
ciency of the cooling system and nDHWS is the efficiency of the DHW
system.
The final energy needs (electricity) for driving the fans of
mechanical ventilation during the heating season were accounted
for as in [20].
For the final energy needs for lighting and appliances, default
annual electricity consumption values were considered (Table 5).
3.7. Thermal solar collection for DHW
The annual energy collected by the thermal solar collector was
estimated as given by Eq. (11) [32].
AESC = [ESC ∗ ASC ∗ nSC ∗ n0 ∗ 365 − a1 ∗ (Tm − Ta )
− a2 ∗ (Tm − Ta )2 ] ∗ 10−3
with ESC the solar irradiation (Wh/m2
day), nSC the solar system effi-
ciency (–), n0 the zero loss efficiency (–), a1 the 1st order heat loss
coefficient (W/◦ C), a2 the 2nd order heat loss coefficient (W/◦ C2 ),
Tm the collectors mean fluid temperature (◦ C), Ta the average ambi-
ent air temperature (◦ C) and ASC the collectors area (m2 ). For the
case-study of Section 4 ahead, the values considered are presented
in Table 6.
The annual energy incident on the solar collector was then esti-
mated using the resource quantification methodology for PV of
Section 3.8 ahead.
3.8. Microgeneration from PV
NZEBs have on-site equipment which “produce” locally energy
carriers that they can export and offset the imports. In practice, the
only energy carrier that is relatively easy to export is electricity,
and the mainstream microgeneration technology is PV panels.
The starting points of the method for quantifying the energy
produced by a PV over a year are the global horizontal and diffuse
771
solar radiation data of climatic files [33]. For each one of the loca-
tions the first step is to compute the declination (ordinal date (N)
Values dependent) [34]. A second step included the calculation of the sun-
0.86 set hour angle and the sunset hour angle for a tilted surface [35].
0.7 Moreover, the ratio of the average daily beam radiation on a tilted
5 surface to that on a horizontal surface for a month Rb , was esti-
0.05 mated by assuming that it has the value which would be obtained
60 ◦ C
if there were no atmosphere [36,37]. The third and final step of the
calculation methodology involves the calculation of the ratio of the
total radiation on the tilted surface to that on the horizontal surface
R, through Eq. (12).
 
ḊH ḊH 1 + cos ˇ 1 − cos ˇ
R = Rb ∗ 1− + ∗ + 0 ∗ (12)
ĠH ĠH 2 2
where 0 is the albedo, ḊH /ĠH is the ratio of the diffuse horizontal
radiation to the global horizontal radiation, and ˇ is the optimum
inclination-slope of the PV module [38]. The yearly solar radiation
incident on the PV is first calculated at an hourly basis, and then
added to result in monthly and yearly totals.
After the previous variables have been calculated, the yearly
electric energy produced by the PV (EPV ) is given by:
(8)
EPV = R ∗ GH ∗ n ∗ APV (13)
(9) with n being the panel efficiency and APV the panel area (m2 ).
The quantification of the electricity produced by a PV system
over a year is used for economic aspects but also to find the size of
(10)
the PV system that is needed for zero yearly balance and thereafter
estimate its initial cost.
3.9. Microgeneration from wind
The quantification methodology for the wind energy supply was
made with hourly wind velocity data and not with yearly data like
in the case of PV and thermal collector, since for this type of equip-
ment that is crucial to ensure a reasonable level of accuracy of the
calculation.
Starting with the wind velocity existing in climatic files (v0 )
(measured at 10 m of height), it is necessary to correct it for the
effects of drag and turbulence caused by ground obstructions. In
order to correct the wind velocity for the height, Eq. (14) [39] was
used:
 h ar
(11) v= ∗ v0 (14)
h0
where h is the height at which the wind velocity v is being
computed-corrected, h0 is the corresponding height (usually 10 m),
of the initial wind speed measurements (v0 ) and ar is a dimension-
less roughness coefficient which depends on the landscape [39].
Moreover, in order to account also for the wind speed reduction
due to turbulence another correction was added, given by [40]:
v1 = v ∗ (1 − 0.01b) (15)
with b the percent turbulence coefficient, which depends on the
site type [40] and v1 the corrected wind speed (m/s).
The effective power P for any given wind speed is given by the
graphs of Fig. 5 for three reference wind turbines, with rated powers
of 400 W, 1.8 kW and 2.2 kW. For the calculation, these curves were
approached to step functions with a width of 1 m/s (Fig. 5).
The corrected wind speed measurements were ‘converted’ to
total annual power output (TPO) for three available types of wind
turbines with a rated respectively, according to Eq. (16), which uses
772 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
Fig. 5. Wind turbine power curves [42–44].
an histogram of the occurrence of each velocity (rounded to the
closest entire value) during the whole year:
28
TPO = nv ∗ P(v)
v=1
where nv is the number of hours per year in which the velocity can
be rounded to the entire value v and P(v) the power output of the
wind turbine when the velocity has the value v.
3.10. Economic indicators
The commonly used indicators for the economic evaluation of
a project, which seem sensible for the purpose of this work, are
the LCC & the Payback Period (PB) methods. For the purpose of
this work, these two “discount” techniques were used in order to
evaluate the economic performance of the design solutions. The
equations used in order to calculate the LCC and the PB are [41]:
N
Cj − Bj
LCC =
j
(1 + d)
j=0
PB
(Bj − Cj )
= c0
j
(1 + d)
j=1
where cj is the cost in year j for the system being evaluated, c0 the
initial cost for the system being evaluated, Bj the benefits (posi-
tive cash flows) in year j for the system being evaluated and d the
discount rate.
3.11. Calculation and optimization software
The equations for characterizing the energy demand-supply
and the economics were implemented in a computer simulation
program developed in Matlab (Version 7.12.0) [42] featuring a
graphical user interface (GUI) for data entry and presentation.
There are two kinds of inputs:
(i) Mandatory: the climate, the geometry of the building, the
energy purchase and sell prices and the intended level of com-
pensation. These inputs define a case-study, since once they
have been inputted they will be treated as fixed during the
simulations;
(ii) Optional: these are the inputs that correspond to the vari-
ables defined in Section 2.3, comprising the insulation level, the
thermal inertia, the infiltration, the ventilation, the window
type, the shading, the glazed area, the orientation, the ther-
mal inertia, the heating system, the cooling system, the DHW
system, the lighting, the appliances, and the microgenera-
(16)
tion systems. If the user selects one of the design alternatives
regarding each of these variables, then it becomes fixed and
the software will only analyze building design alternatives that
consider this option. Otherwise the variable will be treated with
the ‘sweep’ mode, which runs through all the discretized values
of the variable.
The ‘sweep’ feature is an option that is available for all the design
variables that were chosen for the ‘optimization’ part of the tool. For
instance, choosing the ‘sweep’ option in the insulation level input
popup will make the software analyze building design alternatives
with all the discretized values considered. When the ‘sweep’
options is chosen in all the design variables, the outputs such as the
IC and the LCC of all the combinations of design variables are com-
puted.The outputs given by the application for each building design
(17) alternative are the useful heating and cooling needs, the final heat-
ing and cooling needs, the total final energy at which the balance
is met, the area of PV or number of wind turbines needed for the
offset, the annual electricity and/or gas bill the LCC and the IC. They
(18) are exported to a text file every time a simulation is conducted.The
graphical user interface is shown in Fig. 6.
4. Tool application example: low rise building
This section presents a case-study of the use of the tool pre-
sented in the previous section, in order to illustrate the main results
and insights that this methodology can provide. Although the spe-
cific results will be valid for a particular building only, the process
and type of analysis can be replicated for any other building in any
climate.
4.1. Building description
The building chosen to serve as case study is a large one-floor
single-family detached dwelling, with 266 m2 of heated floor area
and five bedrooms (Fig. 7). It was originally designed to be built in
Portugal, and is constituted by 4 bedrooms, a kitchen, 3 bathrooms,
2 offices, a living room, a hall, a corridor and 2 storerooms (Fig. 7).
Table 7 shows the mandatory inputs adopted in the current case
study, which provides also a basic description of the building.
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778 773

Fig. 6. Interface.

Table 8
HDD and CDD.

Climate HDD CDD

Stockholm 4573 36
Lisbon 1190 474
Iraklion 1056 1076

winter but warm summer. The choice of the climates was made
upon consideration regarding the HDD and Cooling Degree Days
(CDD) of each one. Table 8 shows that the HDD of Stockholm are
almost 4 times more than the HDD of Lisbon and Iraklion, while the
CDD range from nearly 0 (Stockholm) to near 500 (Lisbon) and to
over 1000 (Iraklion).
The costs assumed regarding the envelope, the equipment for
the energy services and the renewable energy systems are shown in
Tables 9–11 respectively. They include both the materials and the
labor cost – thus becoming the so-called installed cost. They were
computed based on the market situation of Portugal in 2011. For
Fig. 7. The single-family detached residence. the purpose of this sensitivity study, they were kept equal among
the three climates analyzed (Tables 9–11).
Table 7
Building geometry.
Table 9
Useful area (m2 ) 266 Envelope related costs [44–46].
Shape factor (m2 /m3 ) 0.74
Floor to floor height (m) 2.5 Envelope Costs
Adjacent area (m2 ) 29
Thermal insulation Insulation D/m2 cm 1
Roof area (m2 ) 266
Leakage reduction Attain medium leakage D/m2 13.8
Internal area (m2 ) 18
Attain low leakage D/m2 38.9
Number of dwellings 1
Windows Single glazed clear (Al) D/m2 185
Double glazed clear (Al) D/m2 241
Double glazed clear (Al + TB) D/m2 281
4.2. Climate, cost and energy tariffs scenarios Double glazed low-e (Al + TB) D/m2 316
Triple glazed clear (Al) D/m2 297
Triple glazed clear (Al + TB) D/m2 337
The analysis of the case-study was performed for three climates Triple glazed low-e (Al) D/m2 334
scenarios: Stockholm (Sweden), representing a climate with a cold Triple glazed low-e (Al + TB) D/m2 374
winter, Lisbon (Portugal), representing a climate with a mild win- Shades Cost of shades (inside) D/m2 112
Cost of shades (outside) D/m2 165
ter, Iraklion (Crete, Greece) representing a climate with very mild
774 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778

Fig. 8. Relation between LCC and IC for the climate of Lisbon with hypothetical SP = BP.

Table 10
Equipment for energy services related costs [47].

Equipment for energy services Costs

Heating or cooling equipment Cost of boiler (non-condensing) D/kW 79

Cost of boiler (condensing) D/kW 115
Cost of electric radiator D/kW 19
Cost of RCHP system (conventional) D/kW 310
Cost of RCHP system (class A) D/kW 454
DHW equipment Cost of DHW system (HP) D 693
Cost of DHW system (storage electricity) D 490
Cost of DHW system (storage gas) D 595
Cost of DHW system (tankless condensing) D 1260
Cost of DHW system (tankless gas) D 700
Ventilation Cost of Mechanical Ventilation D/m2 24
Cost of Mechanical Ventilation with Heat Recovery D/m2 48
Lighting Lamps (average efficiency) D 3
Lamps (high efficiency) D 10
Appliances Appliances (average efficiency) D 2892
Appliances (high efficiency) D 6534

Table 11 the combinations of climate scenarios were analyzed considering

Microgeneration costs [48].
only offset from PV modules.
Renewable energy systems Costs

photovoltaic panels PV (average efficiency-12%) D/m2 600 4.3. Results for 3 climates
PV (high efficiency-18%) D/m2 800
DHW equipment Solar collector D 2221 As described in Section 4, setting all design choices to the ‘sweep’
Wind turbines 400 W D 2250
option of the application tool enables the calculation of the initial
1.8 kW D 4900
2.2 kW D 5665
and life cycle costs resulting from all combinations of the design
variables. The number of combinations checked, considering the
14 variables identified in Section 4 is 7,776,000. Fig. 8 shows the
relation between the LCC and the corresponding IC covering all
To focus the analysis of the differences of the results in the cli- the range of possible design solutions for the climate of Lisbon.
matic differences, for the purpose of this study the energy tariffs Each dot in the figure represents a building design solution, i.e.,
were also assumed to be equal among the three climates. A cen- a combination of design variables. To be noted that the IC costs
tral scenario (CS) which has SP = BP (sell price equal to the buy included in the characterization are only those that somehow relate
price, here both taken as 16.5 cents of Euro per kilowatt-hour), was to energy. Figs. 9 and 10 show the results for Iraklion and Stock-
assessed in order to represent the absence of feed-in tariffs, which holm.
is the reality in some countries (e.g. Sweden, Finland, etc.). Even in The first interesting observation, confirmed by Table 12, is that,
the countries with feed-in tariffs, the SP = BP condition could some- for any of the climates, there does not seem to be any trade-off
how be seen as representing a solution of “no cost” to the society. All between LCC and IC, i.e., contrarily to what is frequent, in the

Table 12
Main cost indicators for the SP = BP electricity price scenario (LLCC represents the cost of the design solution with lowest life cycle cost, LIC the cost of the design solution
with lowest initial cost).

IC of LLCC (D/m2 ) IC of LIC (D/m2 ) Percentage LCC of LLCC (D/m2 ) LCC of LIC (D/m2 ) Percentage
difference difference
((IC − LIC)/IC) ((LLCC − LCC)/LLCC)

Lisbon 168 156 7% 221 227 −3%

Iraklion 168 159 5% 224 224 −17%
Stockholm 309 309 0% 380 380 0%
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778 775

Fig. 9. Relation between LCC and IC for the climate of Iraklion with hypothetical SP = BP.

Fig. 10. Relation between LCC and IC for the climate of Stockholm with hypothetical SP = BP.

Fig. 11. Relation between LCC and IC for the climate of Lisbon with hypothetical SP = BP, for the design solutions deviating less than 20% from the lowest LCC.

SP = BP tariff scenario the buildings with better economic perfor-
mance over the life-cycle are trendily also better in terms of initial
cost. This happens because in the NZEB concept buildings that are
more energy-efficient in their operation will need to install less
PV, which tends to be expensive in terms of initial cost. They also
tend to need smaller heating and cooling systems, and therefore
contribute to save in terms of initial cost.
The results further show that for Lisbon the most expensive
NZEB design solution in terms of initial cost is about 3 times more
expensive than the cheapest design solution. The same about 3:1
ratio is observed in terms of life cycle cost. Furthermore, regarding
the climate of Stockholm the same ratios in terms of IC and LCC con-
vert to 7:1 and 6:1 respectively. Recalling that each of the points in
the graph represents a building with zero energy balance over the
year, this clearly illustrates the importance of an economic analysis
at the early design stage.
Another important observation is that the range of initial costs
for Stockholm is considerably higher than that of Lisbon and Irak-
lion. This is attributed to the fact that because of the lower solar
radiation the area of PV necessary to perform the offset is much
higher in Stockholm. This explains the higher ranger of IC observed
in the results (Figs. 8–10 and Table 12).
776 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778

Fig. 12. Relation between LCC and IC for the climate of Iraklion with hypothetical SP = BP, for the design solutions deviating less than 20% from the lowest LCC.

Fig. 13. Relation between LCC and IC for the climate of Stockholm with hypothetical SP = BP, for the design solutions deviating less than 20% from the lowest LCC.

To gain further insights on the ‘near-optimal’ area of the space
of design solutions, Figs. 11–13 show only the design solutions that
deviate at most 20% from the Lowest Life Cycle Cost (LLCC).
Fig. 14 shows the energy density at which the net zero energy
(NZE) balance is achieved for the LLCC and Lowest Initial Cost (LIC)
solutions of the three climate scenarios. The graph shows that the
design solutions for the climates of Lisbon and Iraklion with lowest
LCC and lowest IC are achieved at a much lower energy density than
in the climate scenario of Stockholm. This appears to be more an
expression of the climatic sereneness than of the intrinsic energy-
efficiency of the optimal economic solutions. That can be confirmed
in Table 13, which shows the design solutions of the LLCC solution
in each of the climate scenarios. There it can be confirmed that, for
Lisbon and Iraklion but to some extent also for Stockholm, the cost-
optimal solutions are not at the extreme of the energy efficiency,
but rather that after some point it is more cost-effective to offset
the demand by installing PV than to further decrease it. It must
however be noted that in either case the energy import–export
balance in all climate scenarios is achieved at an energy density
lower than 50 kWh/m2 , a number that accounts for the total final
energy needs including heating, cooling, DHW needs, lighting, and
appliances. Most E.U. countries currently have regulations allowing
about of more than 50 kWh/m2 year just for heating. Therefore
the case-study results show that, without excessively generous

Table 13
Design variable values of the LLCC solution.

Design variables Lisbon Iraklion Stockholm

Insulation 15 cm 8 cm 15 cm
Infiltration Medium Medium Medium
Ventilation Natural Natural Natural
Windows South oriented single glazed clear (Al) South oriented single glazed clear (Al) South oriented triple glazed low-e (Al + TB)
Shading Outside Outside Outside
Glazed area 10% 10% 10%
Heating system RCHP RCHP RCHP class A
Cooling system RCHP RCHP RCHP
DHW system Heat pump Heat pump Heat pump & SC
Lighting High efficiency High efficiency High efficiency
Appliances Average efficiency Average efficiency Average efficiency
PV module High efficiency High efficiency High efficiency
 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
Fig. 14. Energy density at which the NZE balance is achieved for the LLCC and LIC
solutions of the three climates.
feed-in tariffs, an economical-optimal concept of NZEB will push
for more energy efficient buildings and not oppose it as sometimes
feared.
5. Conclusions
The present study presented a methodology and a new tool with
the goal of assisting in the choice of economically efficient NZEB
solutions, right from the early design stage. These can be used for
any residential building in any part of the world, to identify the
LLCC and/or LIC design solutions, considering the local climate, the
endogenous energy resources and the local economic conditions
that lead to a null annual energy balance. This balance was here
made in terms of final energy, although, by using the option of off-
setting only a part of the yearly use, it can be used for cases in which
the designer intends to account for the part of the grid electricity
that is renewable.
In order to obtain indications on the economic space of NZEB
solutions, the relation between the LCC and the corresponding IC
was assessed for a case study of a detached dwelling, covering all
the range of possible design solutions. It was found that as a general
trend the less expensive solutions in terms of life cycle cost are also
less expensive in terms of initial cost. This is somewhat contrary to
the usual expectation that solutions economically more advanta-
geous in terms of life cycle are more costly in terms of initial cost.
The main reason for that seems to be that by investing more on
energy efficiency (which benefits life-cycle costs) it is possible to
install smaller energy services (e.g. HVAC) and micro generation
equipment (PV), and therefore to save initial investment.
Furthermore, as a general trend, the most expensive NZEB
design solution in terms of initial cost was at least 3 times more
expensive than the cheapest design solution. The same at least
about 3:1 ratio, was generally observed in terms of life cycle cost.
This result clearly illustrates the importance of an economic anal-
ysis at the early design stage of NZEBs in order to reach the energy
goals with economic efficiency.
Assessing the influence of different climate contexts in the
economic-efficient design of NZEBs, indicated that the ‘optimal’
design solutions for mild winter climates can be significantly
cheaper than for cold winter climates, as in the latter more energy
efficient solutions are required and also because of the low solar
radiation more PV area is required for the offset which tends to be
more costly. Also, the optimal design solutions for the mild winter
climates are achieved at a much lower energy density than for cold
winter climates.
Though the solutions are more expensive for colder climates,
the importance of an economic optimization is even higher since it
777
was found that the ratio between the worst and the best solution
is about 6:1 whereas for the milder ones it was 3:1 both in terms
of IC and LCC.
It must nevertheless be noted that the tool here presented
intends to provide guidance in the early design stage more than
to replace a detailed dynamic hourly assessment of the solution
that ends up being elected by the designer. In fact, since in parts of
the methodology it was necessary to trade precision for speed, it is
advisable to do so.
References
[1] P. Torcellini, S. Pless, M. Deru, D. Crawley, Zero Energy Buildings: A Critical Look
at the Definition in ACEEE Summer Study, NREL, Pasific Grove, CA, 2006.
[2] J. Laustsen, Energy Efficiency Requirements in Building Codes, Energy Efficiency
Policies for New Buildings, Inernational Energy Agency, 2008.
[3] A. Robert, M. Kummert, Designing net-zero energy buildings for the future
climate, not for the past, Building and Environment 55 (2012) 150–158.
[4] A.J. Marszal, P. Heiselberg, J.S. Bourrelle, E. Musall, K. Voss, I. Sartori, A.
Napolitano, Zero energy building – a review of definitions and calculation
methodologies, Energy and Buildings 43 (2011) 971–979.
[5] I. Sartori, A. Napolitano, K. Voss, Net zero energy buildings: a consistent defi-
nition framework, Energy and Buildings 48 (2012) 220–232.
[6] P. Hernandez, P. Kenny, From net energy to zero energy buildings: defining life
cycle zero energy buildings (LC-ZEB), Energy and Buildings 42 (2010) 815–821.
[7] EU, Directive 2002/91/EC of the European Parliament and of the Council of 16
December 2002 on the energy performance of buildings, Official Journal of the
European Communities (2003) 65–71.
[8] EU, Directive 2010/31/EU of the European Parliament and of the council of 19
May 2010 on the energy performance of buildings (recast), Official Journal of
the European Union (2010) 13–35.
[9] E. Doub, Solar Harvest: City of Boulder’ s First Zero Energy Home, Ecofutures
Building, Inc., Boulder, CO, USA, 2009.
[10] R. Hawkes, Crossway Eco-House, Hawkes Architecture, Staplehurst, Kent, UK,
2009.
[11] K.B. Wittchen, J.S. Østergaard, S. Kamper, L. Kvist, BOLIG+ an energy neutral
multifamily building, in: EuroSun 2010, Graz, Austria, 2010.
[12] M. Kapsalaki, V. Leal, Recent progress on net zero energy buildings, Advances
in Building Energy Research 5 (2011) 123–156.
[13] A. Ferrantea, M.T. Cascella, Zero energy balance and zero on-site CO2 emission
housing development in the Mediterranean climate, Energy and Buildings 43
(8) (2011) 2002–2010.
[14] J. Kurnitski, A. Saari, T. Kalamees, M. Vuolle, J. Niemelä, T. Tark, Cost optimal and
nearly zero (nZEB) energy performance calculations for residential buildings
with REHVA definition for nZEB national implementation, Energy and Buildings
43 (11) (2011) 3279–3288.
[15] M. Hamdy, A. Hasan, K. Siren, Applying a multi-objective optimization approach
for design of low-emission cost-effective dwellings, Building and Environment
46 (1) (2011) 109–123.
[16] A. Hasan, M. Vuolle, K. Sirén, Minimisation of life cycle cost of a detached house
using combined simulation and optimisation, Building and Environment 43
(12) (2008) 2022–2034.
[17] M. Levine, et al., Residential and commercial buildings, in: B. Metz, O.R. David-
son, P.R. Bosch, R. Dave, L.A. Meyer (Eds.), Climate Change 2007: Mitigation,
Contribution of Working Group III to the Fourth Assessment Report of the Inter-
governmental Panel on Climate Change, Cambridge, United Kingdom and New
York, NY, USA, 2007, p. 393.
[18] P. Bertoldi, B. Atanasiou, Electricity consumption and efficiency trends in
European Union, in: JRC Scientific and Technical Reports 2009, European Com-
mission, Joint Research Centre, Institute for Energy, Renewable Energy Unit,
Luxembourg, p. 14.
[19] AECB, AECB Carbon Lite Programme Delivering Buildings With Excellent Energy
and CO2 Performance, AECB, Llandysul, Carmarthenshire, 2007.
[20] Government, Portuguese, Decreto-Lei nr. 80/2006, Regulamento das Carac-
terísticas do ComportamentoTérmico dos Edifícios (RCCTE), Portugal, 2006.
[21] Government, UK, The Building Regulation 2000, Part L: Conversation of Fuel
and Power, C.a.L. Government, NBS, London, UK, 2006, p. 19.
[22] CEN, Energy performance of buildings – calculation of energy use for space
heating and cooling (ISO 13790:2008), Brussels, 2008.
[23] S. Kadam, Zero Net Energy Buildings: Are They Economically Feasible? MIT,
Boston, MA, 2001.
[24] EPA, Energy Star 1992 2010 (cited 2010), Available from: http://www.
energystar.gov/index.cfm?fuseaction=find a product
[25] B. Grinden, N. Feilberg, Residential Monitoring to Decrease Energy Use and
Carbon Emissions in Europe, Analysis of Monitoring Campaign in Europe,
REMODECE, 2006.
[26] A. Gikas, R. Keenan, Statistical Aspects of the Energy Economy in 2004 in Statis-
tics in Focus, Eurostat, Luxemburg, 2006.
[27] M.J.N. Oliveira Panão, S.M.L. Camelo, H.J.P. Gonçalves, Assessment of the Por-
tuguese building thermal code: newly revised requirements for cooling energy
needs used to prevent the overheating of buildings in the summer Energy 36
(5) (2011) 3262–3271.
778 M. Kapsalaki et al. / Energy and Buildings 55 (2012) 765–778
[28] M. Kapsalaki, Economic-efficient Design of Residential Net Zero Energy Build-
ings With Respect to Local Context, Faculty of Engineering University of Porto,
Porto, 2012.
[29] ASHRAE, Non residential heating and cooling load calculations, in: ASHRAE
Handbook Fundamentals, W. Stephen Comstock, Atlanta, 2005.
[30] D. van Dijk, M. Spiekman, P. de Wilde, A monthly method for calculating energy
performance in the context of European building regulations, in: Ninth Inter-
national IBPSA Conference, Building Simulation, Montreal, Canada, 2005.
[31] ESRU, The ESP-r System for Building Energy Simulations, Energy Systems
Research Unit, University of Strathclyde, Glasgow, Scotland, 2002.
[32] G.B.H. Souza, A Multi-objective Decision Support Methodology for Developing
National Energy Efficiency Plans, University of Porto, Faculty of Engineering
University of Porto, 2012.
[33] METEOTEST, Meteonorm 2009, Bern, Switzerland.
[34] P.I. Couper, The absorption of radiation in solar stills, Solar Energy 12 (1969)
333–346.
[35] J.F. Creider, et al., Heating and Cooling of Buildings, McGraw-Hill, New York,
2006.
[36] J.A. Duffie, W.A. Beckman, Solar Engineering of Thermal Processes, Wiley, New
York, 1980.
[37] B.Y.H. Liu, R.C. Jordan, The interrelationship and characteristic distribution of,
direct, diffuse and total solar radiation, Solar Energy 4 (1960) 1–19.
[38] PVGIS, 2008, European Commission, Joint Research Center.
[39] J.F. Manwell, et al., Wind energy Explained-theory Design and Application,
Wiley, New York, 2002.
[40] B. Givoni, Climate Considerations in Building and Urban Design, Wiley, New
York, 1998.
[41] V. Geros, Economic methodologies, in: M. Santamouris (Ed.), Environmental
Design of Urban Buildings: An Integrated Approach, Earthscan, London, UK,
2006.
[42] Mathworks, Matlab, 2011.
[43] Center for Sustainable Building Research, Alliance to Save Energy, and Lawrence
Berkeley National Laboratory, Efficient Windows Collaborative, 1998 (cited
2009), Available from: http://www.efficientwindows.org/
[44] NREL, National Residential Efficiency Measures Database, 2010 (cited 2010),
Available from: http://www.nrel.gov/ap/retrofits/index.cfm
[45] ShadesShuttersBlinds, Shades, Shutters, Blinds, 2001 (cited 2010), Available
from: http://www.shadesshuttersblinds.com/
[46] Canada, National Resources, Retscreen, 1997 (cited 2010), Available from:
http://www.retscreen.net/ang/home.php
[47] Preços, Preços 2004 (cited 2010), Available from: http://www.precos.com.pt/
[48] Energias, Novas, 2010 (cited 2010), Available from: http://novasenergias.
no.sapo.pt/index.htm