Building and Environment 46 (2011) 109e123

Contents lists available at ScienceDirect

Building and Environment

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

Applying a multi-objective optimization approach for Design of low-emission

cost-effective dwellings
Mohamed Hamdy*, Ala Hasan, Kai Siren
Aalto University School of Science and Technology, Department of Energy Technology, P.O. Box 14400, FI-00076 Aalto, Finland

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

Article history: Modern buildings and their HVAC systems are required to be not only energy-efficient but also produce
Received 7 May 2010 fewer economical and environmental impacts while adhering to an ever-increasing demand for better
Received in revised form environment. Research shows that building regulations which depend mainly on building envelope
5 July 2010
requirements do not guarantee the best environmental and economical solutions. In the current study,
Accepted 8 July 2010
a modified multi-objective optimization approach based on Genetic Algorithm is proposed and
combined with IDA ICE (building performance simulation program). The combination is used to mini-
Keywords:
mize the carbon dioxide equivalent (CO2-eq) emissions and the investment cost for a two-storey house
CO2-eq emissions
Investment cost
and its HVAC system. Heating/cooling energy source, heat recovery type, and six building envelope
Summer overheating parameters are considered as design variables. The modified optimization approach performed effi-
Building ciently with the three studied cases, which address different summer overheating levels, and a set of
Optimization optimal combinations (Pareto front) was achieved for each case. It is concluded that: (1) compared with
Pareto front initial design, 32% less CO2-eq emissions and 26% lower investment cost solution could be achieved, (2)
the type of heating energy source has a marked influence on the optimal solutions, (3) the influence of
the external wall, roof, and floor insulation thickness as well as the window U-value on the energy
consumption and thermal comfort level can be reduced into an overall building U-value, (4) to avoid
much of summer overheating, dwellings which have insufficient natural ventilation measures could
require less insulation than the standard (inconsistent with energy saving requirements) and/or addi-
tional cost for shading option.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction in Germany it has lead up to 30% energy savings compared to the

previous standards, in France to 10% savings and in Ireland to
The European Commission announced plans for a European 22e33% savings. Thermal insulation of buildings (external walls,
Union energy policy that included a unilateral 20% reduction roof and floor) and double-pane windows (even triple glazing with
(compared with emission levels of 1990) in greenhouse gas emis- low-e and argon in northern countries like the Baltic States, Finland
sions by 2020. Buildings contribute to about 35% of carbon dioxide and Sweden) reduce annual energy consumption for space heating,
emissions, which is closely related to climate change [1]. The main by lowering heat losses through the building’s envelope [3]. Energy
contributor to the total heating-related CO2 emissions of 725 Mt/ consumption in insulated buildings may be 20e40% less than in
a from the EU building stock in 2002 was the residential sector non-insulated buildings [4].
(77%) while the remaining 23% originated from non-residential Although most of the current European building regulations
buildings. In the residential sector, single-family houses represent have well measures for energy saving, they cannot guarantee the
the largest group and are responsible for 60% of the total CO2 best environmental solutions. Based on models representative for
emissions, equivalent to 435 million ton per year [2]. the range of the Norwegian district heating plants, calculations
Aiming to environmental building solutions, the updated showed that heating-related CO2 emissions in residential buildings
European regulations stipulate well-insulation (U-values) and heat connected to the district heating grid and with an energy standard
recovery requirements. This has resulted in significant energy in accordance with the new building regulations are lower than for
savings for heating, especially in northern Europe: for example, similar buildings with a low-energy standard and with heating
based on electricity [5]. The primary energy use and the CO2
* Corresponding author. Tel.: þ358 9470 23161; fax: þ358 9470 23418. emissions depend strongly on the source of energy supply. A study
E-mail address: [email protected].fi (M. Hamdy). by Gustavsson and Joelsson [6] shows that a single-family house

0360-1323/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.buildenv.2010.07.006
110 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

from the 1970s heated with biomass-based district heating and and ground source heat pump), free cooling option, and three types
with cogeneration has 70% lower operational primary energy use of ventilation heat recovery systems as well as various options for
than if heated with fuel-based electricity. six building envelope parameters: building tightness, insulation
Without considering the economics, the research would lead to thickness of the external wall, roof, and floor as well as the window
costly environmental solutions. For low-emission, cost-effective and shading types.
solutions, Rolfsman’s study [7] addressed two energy saving Since the building energy consumption depends significantly on
measures (e.g., extra insulation and new types of windows) and the desired thermal comfort level, three studied cases are
three heating energy sources (electrical, district heating and proposed. The first case disregards summer overheating, aiming to
ground source heat pump). With the assumed costs, the measures extreme optimal trade-off relation in terms of CO2-eq emission and
did not appear economical. The low-emission cost-effective solu- investment cost. The second case assumes the summer overheating
tion is difficult and time-consuming to achieve using iterative trial level of the non-cooling initial design (building envelope with
and error technique. As an alternative, a suitable simulation-based U-values based on C3-2007 [13]) as a constraint function. The third
optimization approach can be utilized. The increase in the invest- case addresses a higher level of thermal comfort by assuming the
ment cost corresponding to improving the building performance is summer overheating level of the initial design with a cooling option
considered in Hasan et al’s study [8]. The study minimizes the life as a constraint.
cycle cost (LCC) as a single objective, providing one optimal solution
for each studied case. A study by Wang et al. [9] implemented 2. House and its HVAC system
a multi-objective Genetic Algorithm optimization model for single-
storey office building, providing a set of design alternatives for both 2.1. General description
economical and environmental criteria. The study focused only on
the building envelope. However, it was mentioned that more A typical Finnish two-storey house, located in Helsinki, is taken
parameters can be optimized if the scope was expanded to cover as a case study. The gross floor area of the house is 143 m2. The
mechanical systems. Verbeeck and Hens’ study [10] performed life glazing area represents about 15% of the floor area. Each storey has
cycle optimization for externally low-energy dwelling concepts. an internal height of 2.5 m. The two stories are connected by
The study implemented a two-step, multi-objective evolutionary a staircase (Fig. 1a). The internal gains due to people, lighting and
GA algorithm. For a good balance between computational time and electric appliances are assumed according to annual values speci-
approximation of the ideal Pareto front or trade-off curve, 6000 fied by the Finnish building code D5 [14] and taken into the
evaluations were performed. Furthermore, an in-depth analysis calculation as a profile with hourly values.
and fine-tuning was necessary. Legally non-binding building The ventilation system consists of an air handling unit (AHU)
energy codes, voluntary building environmental performance which supplies fresh air to the bedrooms and living room and
assessment schemes, and incentive-based schemes that provide draws the exhaust air from the bathrooms and the kitchen with
subsidies to reduce the costs of improvements are reviewed in a cross air-to-air heat recovery system. The AHU heater keeps the
Ref. [11]. supply air temperature at 18  C when the incoming outdoor air
Expanding the design space solution to include the available temperature is lower than this temperature. The average exhaust
energy sources and ventilation heat recovery types besides the
building envelope parameters has become necessary for environ-
mental building designs. However, the costs should be considered 6.8 m
for economical solutions. The large number of design variables as N

well as the non-linear relations between variables poses a complex This side is adjacent Glassing
challenge for low-emission, cost-efficient building solutions. Bath R. Bed R. to another building area 5 m 2
During the last two decades, evolutionary computation techniques,
such as genetic algorithms (GA), have been receiving increasing
Stair Stair
attention regarding their potential as optimization techniques for
such complex problems. However since GA starts searching by
randomly sampling within an optimization solution space and then
uses stochastic operators to direct a process based on objective
Bed R. Bed R.
function values, a large number of generations are usually required Upper floor
to achieve an acceptable Pareto front. Furthermore, a high quality of
optimal solutions (a continuous Pareto front) cannot be guaranteed
by using a certain number of generations as a stopping criterion
[12].
Living Room
The current study implements a three-phase multi-objective
optimization approach (PR_GA_RF). The approach aims to reduce
the random behavior of the genetic algorithm (GA) by using a good Stair Stair
initial population from the preparation phase (PR). After a low
number of generations, the refine phase (RF) starts using fast and WC Small window for
summer cooling
realistic stopping criteria, considering good diversity for optimal (0.9 m2)
solutions. The approach is combined with IDA ICE 3.0 (Building Kitchen Bed R.
Performance Simulation Program) to minimize the CO2-eq emis- Lower floor
sions related to building energy consumption and the investment
cost of the design variables. A two-storey house in the cold climate House plan Three-zones
of Finland is selected as a case study. (143 m2 floor area) simplified model
In order to achieve environmental cost-effective building solu-
a- Original floor plan b - Three zones simplified model
tions, a wide solution space is suggested, including four heating
energy sources (electrical heating, oil fire boiler, district heating, Fig. 1. Two-storey house (143 m2).
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 111

air flow from the whole house is equal to 0.65 air changes per hour, Table 1
which is in accordance with the Finnish Building code D2 [15]. Design variable.

The building tightness, insulation thickness of the external wall, Variables Type Five initial Lower Upper
roof, and floor as well as window and shading type are assumed as designs bound bound
design variables. An oversized heating system is used in simulation Insulation thickness Continuous 0.124 0.024 0.424
to achieve the minimum acceptable indoor air temperature of external wall [m]
Insulation thickness Continuous 0.21 0.11 0.51
(21 1  C) for different building envelope solutions. No heating/
of roof [m]
cooling system is used for the staircase. Insulation thickness Continuous 0.14 0.04 0.44
of floor [m]
2.2. Initial designs Windows type Discrete 1 1 5
Heat recovery Discrete 2 1 3
Shading type Discrete 1 1 2
Five initial designs are assumed with different heating/cooling Building tightness type Discrete 1 1 5
systems: electrical heating (Sys. 1), oil fire boiler (Sys. 2), district Heating/cooling Discrete From 1 to 5 1 5
heating (Sys. 3), ground source heat pump GSHP (Sys. 4), and GSHP system type
with a free cooling option (Sys. 5). The cooling option is assumed
only with the 5th initial design. The heat recovery has an annual
efficiency of 70%. Based on the National Building Code of Finland, Three different ventilation units are presented in Table 3. Table 4
C3-2007 [13], U-values 0.24, 0.15, 0.24, and 1.4 W/m2 K are assumed presents the shading options: external blind, horizontal laths
for external walls, roof, ground floor, and window respectively. The shading and no shading option. Improving the building tightness is
window type has a shading coefficient SC of 0.656. The initial done by careful work and more strict control on the site.
solutions are assumed with a shading option and building tightness This creates an additional cost. It is assumed that the tightness
n50 ¼ 4 1/h where n50 is the number of air changes per hour n50 ¼ 4 1/h is the reference value with no additional costs. A smaller
equivalent to the air leakage rate with a 50 Pa pressure difference tightness value creating an additional cost (V/floor m2) is shown in
between indoor and outdoor. Table 5. Five types of heating/cooling systems with different
emission factors are shown in Table 6: four heating systems (direct
2.3. Simplified model including operable window electrical heating, fuel oil boiler, district heating, or GSHP) and only
one heating/cooling option (GSHP with the free cooling option).
To reduce the execution time of simulation, a simplified model The ground source heat pump (GSHP) is a brine-to-water system
(Fig. 1b) is used to represent the house by three zones: lower floor, with a compressor driven by electricity. The brine flows in a deep
upper floor, and staircase. The zones are considered as open plan borehole using the ground as the heat source. Cooling is achieved
areas. On the north and south directions, two windows are used at without compressor, while the compressor is kept in the free
each storey to represent the total fixed glazing area of the building cooling mode, by cold water which flows between the chilled floors
(20 m2). In addition, a small window (0.9 m2) is proposed and and an intermediate heat exchanger. The free cooling upgrade costs
installed at the middle of each storey with a special PI-controller. an additional 1000 V. However, it provides powerful cooling, using
The two small windows open proportionally when the outdoor air negligible amount of energy to run a circulating pump.
temperature is less than the indoor air temperature and the latter is
higher than 24  C. The controller tries to emulate human behavior 3.2. Objective functions
by opening the window to improve the thermal comfort by
increasing the ventilation in summer. The aim of this study is to achieve low-emission, cost-effective
design solutions. Therefore CO2-eq emissions related to heating
3. Formulation of the optimization problem energy and the investment cost related to the suggested design
variables are selected as two objective functions to be minimized.
3.1. Design variables The first objective, CO2-eq emissions [kg/m2 a], is calculated by the
following equation.
Aiming to low-energy cost-effective concepts for single-family
houses in Finland, a sensitivity analysis was performed [16] CO2 -eq ¼ Q  EF=h (1)
addressing insulation thickness, window type, building tightness,
and efficiency of heat recovery as design parameters. In a recent where Q is the total heating energy, EF is the primary greenhouse
Finnish optimization study [8], nearly the same parameters with gas emission factor, and h is the heating system efficiency (Sys. 1e3)
different assumptions were addressed, and the life cycle cost was or the COP of the heat pump (Sys. 4 and 5), see Table 6.
considered as a single-objective function. Different types of energy The second objective, the investment cost, is the total cost of the
sources are used in Finland. The primary greenhouse gas emission investments related to the eight design variables. Different types
factors (related to three major greenhouse gases: CO2, sulphur, and of insulation (mineral wool, blow-in wool, and polyurethane) are
nitrogen.) for different types of energy supplied to the buildings used in the external wall, roof, and ground floor. The insulations
were evaluated in Ref. [17]. have prices of 56.3, 32.5, and 100 V/m3, and thermal conductivity of
The current study considers eight design variables: heating/
cooling energy source, ventilation heat recovery type, building Table 2
tightness, window type, shading option as well as insulation Window type.
thickness of the external wall, roof, and floor. The environmental
Window U-value Total solar Shading coefficient Price
impact is estimated considering emission factors from Ref. [17]. The Type [W/m2 K] transmission [%] factor [V/m2]
eight design variables and their investment costs are presented in 1 1.4 44 0.656 180
Tables 1e6. Table 1 presents the initial values, lower bounds, upper 2 1.1 44 0.656 185
bounds, and types (discrete or continuous) of the eight design 3 1 34 0.53 205
variables as well as the ranges of the external wall, roof, and floor 4 0.85 29 0.482 240
5 1.1 28 0.437 210
insulation thickness. Five window types are described in Table 2.
112 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

Table 3 Table 5
Price of the AHU including different types of heat recovery system. Building tightness type.

Type Efficiency [%] Specification Price [V] Type Specification Price for additional
1 60 Plate heat exchanger 3172 n50 [l/h] work [V/m2]
2 70 Rotating wheel 3443 1 4 0
3 80 Rotating wheel 3715 2 3 5
3 2 12
4 1 22
0.035, 0.05, and 0.026 W/m2 K, respectively [8]. Only one shading 5 0.5 30
option (External blind, horizontal laths) is assumed with the price
from [18]. The investment costs of other design variables are
assumed to be in line with market prices and presented in Tables 2, Method: a modified multi-objective optimization approach
3, 5, and 6. (PR_GA_RF). (More details about this approach can be found in the
following section).
3.3. Constraint function
4. The modified optimization approach
The HVAC energy consumption (consequently the CO2-eq emis-
sions) depends on the criterion of thermal comfort, i.e., disregarding In a previous study [12], two combinations between the opti-
the summer overheating allows additional insulation for maximum mization algorithms were developed. The First is called PR_GA. It is
space heat energy saving. To assess the level of summer overheating, a two-phase multi-objective optimization combination that works
degree-hour DH24 is used and defined as the summation of the under the MATLAB environment. Briefly, in PR_GA the Genetic
operative temperature degrees higher than 24  C at the warmest Algorithm (GA) from the MATLAB 2008a Genetic and Direct Search
zone during a one-year simulation period (8760 h), as follows: Toolbox [19] was modified to be able to deal with discrete and
continuous variables. Then it was combined with a deterministic
8760
i ¼X
optimization solver (FMINCON, single-objective deterministic
DH24 ¼ dH24 (2)
solver from MATLAB optimization ToolBox) in order to provide GA
i¼1
with a good collection of individuals as an initial population. This
process is called the preparation phase (PR). The major advantage of
dH24 ¼ ðTi 24ÞDt when Ti 24 > 0
PR_GA is that it tries to reduce the random behavior of GA in an
attempt to obtain good solutions with a lower number of evalua-
dH24 ¼ 0 when Ti 24  0 tions (simulation runs).
where Ti is the operative temperature [ C] at the centre of the The second combination is called GA_RF. It can be used for high
warmest zone (upper floor) and dt is a one hour time period [h]. quality, enhanced optimal solutions. As a hybrid optimization
In order to investigate the influence of thermal overheating on solver, GA_RF utilizes Fgoalattain (Multi-objective optimization
the optimal solutions, three case studies (Cases 1, 2 and 3) are solver from MATLAB 2008a optimization toolbox) after using the
proposed. Case 1 disregards the summer overheating, aiming to modified GA (MATLAB GA able to deal with discrete and continuous
extreme optimal trade-off relation between the objectives. In Case design variables) in order to improve the quality and diversity of
2, the degree-hours DH24 value of the initial design without obtained results. The advantage here is that GA_RF refines only
a cooling option (DH24 ¼ 2400  C h) is considered as a constraint a certain number of the obtained solutions exhibiting significant
function. For a higher thermal comfort level, DH24 ¼ 1000  C h (the diversity. This reduces the time of the optimization process by
indoor degree-hours of the initial design with a cooling option) is avoiding much of mostly similar evaluations. The other advantage
assumed as a constraint function for Case 3. The three studied cases is that the Pareto solutions are enhanced by sorting the history-
can be summarized as follows: results of the refining steps.
Objective 1: Minimum CO2-eq emissions related to heating The current study proposes a combination between the two
energy. predefined approaches. The new combination is called PR_GA_RF.
Objective 2: Minimum investment costs for the eight design This combination provides the advantages of the two modified
variables. approaches (PR_GA and GA_RF). However, since the PR_GA_RF is
Constraint: a multi-step approach, many of the evaluations are repeated. This
could be avoided easily if the new combination (PR_GA_RF) was
Case 1 (no constraint); modified to determine the values of the objective and constraint
Case 2 (DH24  2400  C h); functions of the repeated design, i.e., variable combinations directly
Case 3 (DH24  1000  C h). from the history data.

Find: optimal combinations between the design variables (Xi) Table 6

where i ¼ 1e8. Heating/cooling system types.

Type Type Price [V/m2] EF/h [kg CO2-eq

Table 4 /kWh]
Shading type. 1 Direct electric radiatora 30 0.459/1
2 Oil fire boilera 94 0.267/0.9
Type MSC MSSC Description Price [V/m2] 3 District heatinga 101 0.226/1
1 0.14 0.09 External blind, 200 4 GSHPa 126 0.459/3b
horizontal laths 5 GSHP with free cooling 133 0.459/3b
2 1 1 No shading 0
EF: emission factor.
a
MSC: multiplier for shading coefficient; MSSC: multiplier for solar shading Without cooling system.
b
coefficient. h ¼ COP in case of GSHP system.
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 113

MATLAB Environment

Write Input file Read IDA ICE

Optimizer (IDALISP) (Building
(Modified performance
MATLAB Simulation
optimization CO2-eq, Outputs files program)
approach) Investment cost, Read (energy consumption,
Write
and DH24 indoor temperature)
calculator
(MATLAB m-files)

Data source
(emission factors
and prices)

Fig. 2. Components and their relationships for simulation-based optimization.

In the current study, the preparation phase (PR) used FMINCON consumption, etc. The accuracy of IDA ICE was assessed through the
to minimize the CO2-equivent emissions then the investment cost IEA solar heating and cooling program, Task 22, subtask C [21].
as single objectives in two separate optimization steps imple- IDA ICE 3.0 was chosen as one of the major 20 building energy
menting a constraint function on the degree-hour DH24 for Cases 2 simulationprograms, which were subjected to analysis and
and 3. The history of the PR’s iterations is sorted based on the two comparison [22].
objective functions providing the GA with a good initial population. The combination between PR_GA_RF and IDA ICE was used to
The refine phase (RF) addressed only 10 optimal solutions from the perform the optimization process for the three cases studied. The
GA Pareto front. The refined optimal solutions were selected work was done by a computer (IntelÒ core Ô2 Quad CUP 2.40 GHz
considering different investment costs dividing the Pareto front processor. 3061 MB RAM) with the Windows Vista system.
with fixed step (maximum investment cost minimum invest- A simulation run took on average 2.5 min.
ment cost)/10. In order to obtain the final Pareto front, the history of
the refine phase (RF) is sorted applying the non-dominated
6. Results and discussion
concept.

6.1. Case 1
5. The simulation-optimization approach
In this case the optimization approach is employed to obtain an
PR_GA_RF is combined with IDA ICE (IDA Indoor Climate and extremely optimal relation between the two objective functions.
Energy 3.0 program [20]) under the MATLAB environment as No constraint function is assumed on summer overheating, which
shown in Fig. 2. IDA ICE is a whole-building dynamic simulation provides possibilities for more energy saving, e.g., using more cost-
program that makes simultaneous performance assessments of all effective insulation to reduce the space heating energy. For this
issues fundamental to building design: shape, envelope, glazing, purpose, 1310 simulation runs are performed: 290 for the prepa-
HVAC systems, controls, light, indoor air quality, comfort, energy ration phase (PR), 720 for the GA phase using 40 population

Oil fire boiler (Sys.2)

GSHP District Electrical radiator
(Sys.4) (Sys.3) (Sys.1)
40
Cadidate solutions
Investment cost of the eight

Pareto front
35
design variables [103 €]

Initial designs
Utopia point
30

25

20
Utopia point
15

10
10 20 30 40 50 60 70 80 90 100
CO2-eq emission [kg/m2 a]
Fig. 3. Non-cooling initial designs and the Pareto front solution of Case 1 (DH24  6250  C h).
114 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

individuals (Pop) and 18 generations (Gen), and 300 for the refine 3, and 4) giving low-energy cost-effective solutions. In the optimal
phase (RG). Fig. 3 shows all the candidate solutions from the opti- solutions of Fig. 3, the space heating energy (AHU heating coil plus
mization history, four non-cooling initial designs (Table 1: heating space heating device energy) comes in the descending order from
system from 1 to 4), 41 optimal solutions (Pareto front), and the right to left, and as shown in Fig. 4. It should be noted that this is
utopia point (minimum investment cost and minimum CO2-eq achieved by finding cost-effective combinations between building
emissions). The values of the objective functions corresponding to envelope parameters to get a lower building U-value (Ubldg). The
the utopia point are calculated by assuming the lowest price for Ubldg represents the mean U-value for the whole-building envelope
each design variable (lower bounds of insulation thickness, defined by the following equation:
window type 1, non-shading option, building tightness n50 ¼ 4 l/h,
the lowest efficient heat recovery type 1, and electrical radiation as 
a heating system) to estimate the minimum investment cost, and Ubldg ¼ Uwall  Awall þ Uroof  Aroof þ Ufloor
by assuming the best heating energy saving measures (upper .
bounds of insulation thickness, window type 4, non-shading Afloor þ Uwindow  Awindow Atotal ð3Þ
option, building tightness n50 ¼ 0.5 1/h, and the highest efficient
heat recovery) as well as GSHP as a heating system to determine the where
minimum CO2-eq emissions. The utopia point represents an
Atotal ¼ Awall þ Aroof þ Afloor þ Awindow
imaginary target which cannot be achieved because of the trade-off
between the objectives. The utopia point is used here to check the From Fig. 4, it can be seen that, for each system, lower Ubldg
extremes of the obtained Pareto front. provides lower space heating energy. The variations in the selected
The cooling option is only available in Sys. 5 (GSHP with types of heat recovery and building tightness (Fig. 5) explain the
a cooling option). This requires additional investment cost (1000 V) small deviation in the relation between the results of Ubldg and
compared with Sys. 4 (GSHP without a cooling option). Since this space heating in Fig. 4. The average Ubldg of the optimal solutions is
case is performed without a constraint on summer overheating 0.238 W/m2 K, and about 80% of the solutions have Ubldg less than
level (DH24), there is no reason for additional investment in that for the initial design (Ubldg ¼ 0.3 W/m2 K). The minimum
a cooling option. Therefore Sys. 5 is not selected in any of the 41 bound for Ubldg in the solution space is 0.131 W/m2 K (solution 1).
optimal combinations. The same reason applies for not getting However for Sys. 1, reducing Ubldg to less than 0.17 W/m2 K (solution
solutions with window shading. 23) was not an optimal solution because a lower emission heating
Disregarding the summer overheating provides a set of solutions system (oil fire boiler) was available. Fig. 4 indicates that the
(Pareto front) which are better than the initial ones in terms of lower optimization solver changed the heating system from Sys. 1 (solu-
CO2-eq emissions and investment cost. However, the obtained tion 23, Ubldg ¼ 0.17 W/m2 K) to Sys. 2 (solution 22, Ubldg ¼ 0.34 W/
solutions have higher DH24 levels (DH24 from 4181 to 6254  C h) m2 K) instead of continuing in reducing Ubldg with Sys. 1 from
than the non-cooling initial ones (DH24 ¼ 2400  C h). Also, it is solutions 41 to 23. In addition, at solution (23), building tightness
worthwhile to mention that the obtained solutions on the Pareto (n50 ¼ 1 1/h) and heat recovery efficiency (h ¼ 70%) were selected,
front are found to be classified according to the heating energy while the solution space includes higher building tightness
source. This reveals that the heating system has a stronger influence (n50 ¼ 0.5 1/h) as well as better heat recovery (h ¼ 80%), which
on the low-emission cost-effective solutions than the other design provides higher savings in the space heating energy. This illustrates
variables. This phenomenon can be explained by the following: that changing the heating system type from Sys. 1 to Sys. 2 is
a better solution for the two objective functions than continuing in
1. The amount of CO2-eq emissions is the product of the emission reducing the space heating energy with Sys. 1. It is also to be noted
factor (EF) and the total heating energy (Q), see Equation (1). that the district heating solutions (Sys. 3) dominated the oil fire
The emission factor (EF) depends only on the type of the boiler solutions (Sys. 2) except for one solution. The reason is that
heating source, see Table 6. while the emission factor of Sys. 3 is less than that for Sys. 2 there is
2. The total heating energy (Q) consists of space heating (qs), no considerable difference in the price between the two systems.
domestic hot water heating (qw), and system heat loss (ql). The Compared with the initial design (Sys. 1), 32% lower CO2-eq
domestic water heating and system heat loss are assumed emissions and 26% lower investment are achieved on the Pareto
constants. Their related CO2-eq emissions (EF  qw and EF  ql) front. The minimum required air temperature (21 1  C) inside the
are functions of the emission factor (EF) of the heating system. upper and lower storey for all the optimal building envelope
Based on the 41 obtained solutions (Pareto front), the domestic solutions is achieved. A procedure is made available to open a small
hot water heating together with the system heat loss is window at each storey via a PI-controller when the indoor air
(36.5 kWh/m2 a), which is 25e52% of the total heating energy. temperature is >24  C and the outdoor air temperature is lower
3. The investment cost of the heating system is often higher than than the indoor air temperature. However, the indoor operative
the cost of the other design variables. According to the temperature is raised above 25  C for 13% of the year hours with
41 obtained solutions, the price of an electrical radiator heating a maximum temperature of 33.2  C. This minimum overheating
system (the cheapest heating system) and the price of a Ground occurred at solution number 41, which has the maximum value of
Source Heat Pump (the most expensive heating system) Ubldg (0.496 W/m2 K) in the optimal solution and the maximum
represent 11% and 50% of the maximum investment cost CO2-eq emission (70.5 kg CO2-eq emission/m2). The next two
(38 500 V), respectively. studied cases minimize the CO2-eq emissions and investment cost
while trying to avoid much of summer overheating. Section 8
It is worthwhile also to mention that without changing the compares between the three cases presenting the duration curves
heating system, there is only one way to reduce the CO2-eq emis- of the operative temperature and mean air temperature for them.
sion, namely by reducing the energy consumption (Q). Since the
domestic hot water heating (qw) and system heat loss (ql) are 6.2. Case 2 (DH24  2400  C h)
assumed constants, optimal combinations between the insulation
thicknesses, window type, building tightness, and heat recovery DH24 ¼ 2400  C h is the summer overheating level of the four
system are found by the optimizer for each heating system (Sys. 1, 2, non-cooling initial designs (initial design values in Table 1, heating
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 115

No shading

Oil fire boiler (Sys.2)

GSHP District Electrical radiator
125

Space heating energy [kWh/m2.a]

(Sys.4) (Sys.3) (Sys.1)

100
Initial designs 91.5 kWh/m2.a

75

50

25

0
0.5

0.4
U-Value bldg [W/m2.K]

Initial designs
0.3
0.3 W/m2 K

0.2

0.1

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Optimal Solutions

Fig. 4. Space heating energy of the optimal design as a percentage of the space heating energy of the initial design.

systems from 1 to 4). In order to achieve low-emission cost-effec- is limited) for the initial design (Table 1) which has a poor building
tive solutions at that level of summer overheating, DH24  2400  C tightness (n50 ¼ 4 1/h) and shading option (Table 4). This means
is implemented as a constraint function. In this case, the modified that additional insulation, higher building tightness, and/or non-
multi-objective optimization approach PR-GA-RF (Section 4) per- shading solutions may cause summer overheating even if the
formed 2787 evaluations: 297 for the preparation phase (PR), 1200 cooling option is selected. The challenge was to find optimal
for GA phase, and 1290 for refine phase (RF) achieving 64 optimal combinations of building envelope parameters which achieve the
solutions. Fig. 6 shows those solutions in addition to the four non- predefined thermal comfort criterion (DH24  2400  C h) with
cooling initial designs. a smaller increase in energy (consequently a smaller increase in the
To check the extremes of the obtained Pareto front, the utopia CO2-eq emissions) and a smaller increase in the investment cost
point is determined as shown in Fig. 6. Because of the summer (e.g., no mechanical cooling solutions).
overheating limit of Case 2 the values of the objective functions Fig. 7 shows that the optimization solver kept the
corresponding to the utopia point are found by applying a modified DH24  2400  C h for all the 64 optimal solutions, and half of those
single-objective optimization solver (Fmincon solver, from the solutions are without a cooling system. This is achieved by using
MATLAB 2008a optimization toolbox, modified to deal with the shading option for all the non-cooling solutions (33e64). The
discrete and continuous variables) to determine the minimum shading option was also used for the free-cooling ones (1e9) when
amount of CO2-eq emissions then the minimum investment cost at implementing additional insulation (Ubldg  0.17 W/m2 K) for more
two separate steps using DH24  2400  C h as a non-linear space heating energy saving. For each heating/cooling system,
constraint function. According to the utopia point in Fig. 6, the PR- optimal combinations between the window type and insulation
GA-RF optimization approach achieved the extreme minimums of thicknesses of the external wall, floor, and ceiling were selected,
the two objectives separately at the two terminals of the Pareto making a clear relation between the overall heat transmission
front. It is also worthwhile to mention that the optimal solutions of coefficient Ubldg and energy consumption, as shown in Fig. 8. The
this case are found to be classified according to the type of a heating influence of the other design parameters is causing some irregu-
system similar to Case 1. However system 5 (GSHP with cooling larity in the relation.
option) is selected in 32 optimal solutions. The building envelope’s discrete variables played a big role in
System 5 is the only system in the solution space that offers attaining the defined level of the thermal comfort
a cooling option. Furthermore, it has the highest investment cost. (DH24  2400  C h), with and without a cooling option. For
The cooling system is sized (i.e., the potential of free cooling system example, to avoid much of summer overheating, the building
116 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

No shading
Oil fire boiler (Sys.2)

GSHP District Electrical radiator

(Sys.4) (Sys.3) (Sys.1)
Building Tightness T ype 5

0
3
Heat Recovery T ype

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41
Optimal Solutions
Fig. 5. Average U-value of the building (Ubldg) for the optimal solutions of Case 1 (DH24  6250  C h).

Initial dsigns (DH24 = 2400 C.h)

Pareto front (DH24 ≤ 2400 C.h)
Utopia point (DH24 = 2400 C.h at minimum CO2-eq emissions)

Solutions Solutions Solutions Solutions Solutions

1:9 10:32 33:34 35:45 46:64

45 GSHP without
Investment cost of the eight

cooling (Sys. 4)
5
design variables [103 €]

District heating
40
(Sys. 3)
Oil fire
35 boiler
(Sys.2)
30 Electrical Radiator(Sys. 1)

25
GSHP (Sys.5) District
Oil fire boiler
heating
20 (Sys.2)
(Sys.3)
Utopia point
15
10 20 30 40 50 60 70 80 90 100
2
CO2-eq emission [kg/m .a]

Fig. 6. Non-cooling initial designs and the Pareto front of Case 2 (DH24  2400  C h).
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 117

District heating
With shading Without shading (Sys.3) With shading

GSHP with free cooling Oil fire boiler Electrical Radiator

(Sys.5) (Sys.2) (Sys.1)
2400

DH 24 [C .h]

1600

800

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64
Optimal Solutions

Fig. 7. DH24 for the 64 optimal solutions of Case 2 (DH24  2400  C h).

tightness type 1 (n50 ¼ 4 1/h) and the shading option are selected (optimal solution from 64 to 46) with incrementally additional
for 80% and 66% of the optimal solutions, respectively and as shown insulation to minimize the space heating energy (Fig. 8), and
by Fig. 9. In addition, window types 1 and 2, which have the highest consequently CO2-eq emissions (Fig. 6). However, since any addi-
shortwave shading coefficient (SSC ¼ 0.656: less solar heating tional insulation would increase the summer overheating, the
energy), are selected for 88% of the optimal solutions. Low expenses optimization solver switched to the Oil Fire Boiling heating system
of these combinations are the other reasons behind the selection. (Sys. 2) at solution 45, avoiding any increase in the building enve-
Next we analyze the performance of the optimization solver in lope insulation (Fig. 8). This provides a higher reduction in CO2-eq
selecting various design parameters. To obtain minimum solutions emissions (Fig. 6), because Sys. 2 has a lower emission factor than
of investment cost and CO2-eq emissions, the Electrical Radiator Sys. 1 (Table 6) and keeps DH24 below 2400  C h as shown in Fig. 7.
(Sys. 1), the lower price heating system, is found to be optimal This requires a significant increase in the investment cost (the

District heating (Sys.3)

With shading Without shading With shading

GSHP with free cooling Oil fire boiler Electrical Radiator

175
Space heating energy [kWh/m2.a]

(Sys.5) (Sys.2) (Sys.1)

150

125

100
Initial designs 91.5 kWh/m2.a
75

50

25

0.7
0.6
U-Value bldg [W/m2.K]

0.5
0.4

0.3 Initial designs

0.3 W/m2 K
0.2
0.1
0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64

Optimal Solutions

Fig. 8. Ubldg and the space heating energy of the optimal solutions of case 2 (DH24  2400  C h).
118 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

District heating
With shading Without shading (Sys.3)
With shading

GSHP with free cooling Oil fire boiler Electrical Radiator

(Sys.5) (Sys.2) (Sys.1)
5
Building Tightness Type
4

0
Window Type

0
3
Heat Recovery Type

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64

Optimal Solutions

Fig. 9. Building tightness, window, and heat recovery type of the optimal solutions for Case 2 (DH24  2400  C h).

difference between the prices of Sys. 2 and Sys. 1), explaining lower heating energy solutions: from 46 to 50 for the Electrical
the discontinuity (investment cost step) in the Pareto front at Radiator (Sys. 1) and from 1 to 23 for GSHP (Sys. 5) optimal solu-
8620 kg/a of CO2-eq emissions as exhibited in Fig. 6. A higher tions. This provides lower CO2-eq emission solutions, while any
efficiency heat recovery types 2 and 3 (Table 3) are selected for additional insulation is avoided to keep DH24 within an acceptable

36
Pareto front (DH≤ 1000 C. h)
Investment cost of the eight

35 Initial design (DH = 1000 C. h)

design variables [103 €]

34

33 Alternative
Solutions
32

31 Better
Solutions
30
16 18 20 22 24 26 28

CO2-eq emission [kg/m2.a]

Fig. 10. Initial designs and the Pareto front of Case 3 (DH24  1000  C h).
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 119

Sys.5 (GSHP with free cooling )

Building tightness type 1(n50 = 4 1/h)

With shading
2
Ubldg < 0.3 W/m K Ubldg ≥ 0.3 W/m2 K

Alternative Better Solutions Alternative

Solutions Solutions
1000

800
DH 24 [C. h]

600

400

200

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58

Optimal Solutions

Fig. 11. DH24 for the optimal solutions of Case 3 (DH24  1000  C h).

Sys. 5 (GSHP with free cooling )

Building tightness type 1(n50 = 4 1/h)

With shading
Ubldg < 0.3 W/m2 K Ubldg ≥ 0.3 W/m2 K

Alternative Better Solutions Alternative

Space heating energy [kWh/m2.a]

150
Solutions Solutions
125

100
Initial designs 91.5
kWh/m2.a
75

50

25

0.6

0.5
U-Value bldg [W/m2.K]

0.4
Initial designs
0.3
0.3 W/m2 K
0.2

0.1

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58
Optimal Solutions

Fig. 12. Ubldg and the space heating energy of the optimal solutions for Case 3 (DH24  1000  C h).
120 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

Sys.5 (GSHP with free cooling )

Building tightness type 1(n50 = 4 1/h)

With shading
Ubldg < 0.3 W/m2 K Ubldg ≥ 0.3 W/m2 K

Alternative Better Solutions Alternative

Solutions Solutions
5

4
Window Type

0
3
Heat Recovery Type

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58
Optimal Solutions

Fig. 13. Window and heat recovery type of the optimal solutions for Case 3 (DH24  1000  C h).

limit (Fig. 7). Since the prices of the Oil Fire Boiling (Sys. 2) and maximum temperature. For a higher thermal comfort level,
District heating system (Sys. 3) are close, the cheapest heat DH24  1000 is assumed as a constraint function in Case 3. Since the
recovery system (type 1) is selected for all the optimal solutions of major part of the solution space does not offer any cooling option,
those two systems, giving them the opportunity to compete with a larger number of evaluations, i.e., 3500 (325 PR þ 40 Pop X 40
each other. Since the Oil Fire Boiler (Sys. 2) has a lower price, it Gen þ 1575 RF), was needed to achieve high level thermal-comfort
dominates most of the District Heating (Sys. 3) solutions. On the solutions using strictly the constraint function (DH24  1000  C).
other hand, the GSHP Free-Cooling optimal solutions (32 to 10) Fig. 10 exhibits 58 optimal solutions for this case representing the
dominate another part of the District Heating solutions saving trade-off relation between the two objectives. Those solutions are
much of the investment cost by selecting the non-shading option. compared with the initial design (Table 1, Sys. 5 GSHP with free
However, the optimal solutions from 1 to 9, in addition to using cooling option) which has DH24 of 1000  C h. It is found that only three
extreme insulation (Ubldg  0.17 W/m2 K) for the lowest CO2-eq optimal solutions provide better designs than the initial one; the
emissions, use the shading option to maintain the thermal comfort other 55 solutions can be considered as alternatives.
criterion. Since some amount of overheating is allowed, the shading
Sys. 4 has the same emission factor (EF) as Sys. 5 (GSHP with free option 1 (less direct solar heat gain) and the leakiest building type
cooling option). However to maintain the DH24 constraint, a high (n50 ¼ 4 1/h) in addition to Sys. 5 (the only system which has
expense shading option is necessarily for Sys. 4 (GSHP without cooling option) are selected for all the optimal solutions. This keeps
a cooling option) solutions. Since the shading option is more the summer overheating under the desired threshold
expensive than the free cooling option of Sys. 5, all Sys. 4 solutions (DH24 ¼ 1000), as shown in Fig. 11.
are dominated by Sys. 5 ones. This explains the absence of Sys. 4 The optimization solver tried to minimize the CO2-eq emissions
among the optimal solutions. For the same reason, the Sys. 5 using optimal combinations between the building envelope param-
optimal solutions (1e32) dominate the initial designs of Sys. 4 eters and heat recovery type. The Ubldg of those combinations is
(Fig. 6). arranged in a descending order from right to left (Fig. 12). The initial
design has Ubldg of 0.3 W/m2 K. This value is used to classify the
6.3. Case 3 (DH24  1000  C h) optimal solution as shown in the following.
Fig. 13 shows that for optimal solutions 27e58, window types 1
In the previous section, DH24 2400  C h is assumed as a constraint and 2 and heat recovery type 1 (low price combination) are selected
function. This leaded to thin the operative temperature exceeds 24 with Ubldg  0.3 W/m2 K (less insulation cost) giving a lower
and 25  C for 18.5 and 8% of the year hours, respectively, with 30.5  C investment cost than the initial design does. A higher efficiency
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 121

Fig. 14. Brute force, feasible solutions, and Pareto front for case 3 (DH24  1000  C h).

heat recovery (type 2) is utilized to obtain a set of solutions (21e26) Case 3. Brute-force search is an exhaustive search that systematically
which produce smaller amounts of CO2-eq emissions (Fig. 10) enumerates all possible candidate solutions.
without an increase in summer overheating (Fig. 11). For more The predefined case study includes eight design variables. Three
reduction in CO2-eq emissions, more insulation is used of them are continuous and five are discrete. Table 1 presents the
(Ubldg < 0.3 W/m2 K). Moreover, window type 1 (SSC ¼ 0.656) then upper and lower bounds of the design variables. If 5 cm
window type 5 (SSC ¼ 0.437) are selected to reduce the space is considered as an exhaustive-search step for the thickness of
heating demand by increasing the direct and indirect solar heat insulation in the wall, roof, and floor, and all combinations
gain respectively. The influence of the U-values of those window between the discrete variables are taken into account, then
types 1 and 5 (1.4 and 1.1 W/m2 K) is shown in Fig. 12 where the 8  8  8  5  3  2  5  5 ¼ 384 000 simulation runs are needed
resistance of the insulation thicknesses is nearly constant (optimal to get all possible candidate solutions. The execution time of one
solutions from 1 to 20). The lower U-values of window types 2 and simulation run is about 2.5 min. This means that 666 days would be
4 (1 and 0.85 W/m2 K respectively) are not used, thus avoiding required to get the brute-force search results for the predefined
additional overheating in order to keep the building at problem. To make the brute-force search feasible, the problem size
DH24  1000  C h as defined by the constraint function. The highest needs to be limited by using problem specifying heuristics, which
efficiency heat recovery type 3 is used, providing more reduction in help to reduce the number of candidate solutions to a manageable
the heating energy (consequently CO2-eq emissions) without size. We can get some indications from the PR_GA_RF results that
increasing the summer overheating. This is shown by the optimal will make the brute-force search fall within a preferable set of
solutions (Ubldg < 0.3 W/m2 K) in Fig. 13. solutions. For example, in Case 3 the heating/cooling system 5,
shading option, and building tightness type 4 are selected in all the
optimized solutions. However, window types 3 and 4 are not
7. Comparison with results from a brute-force search method selected (Fig. 13). This can reduce the brute-force candidate solu-
tions significantly. More reduction is obtained by using a larger step
A brute-force search method is implemented to check the results for the continuous variable (8 cm instead of 5 cm for the
obtained by the modified optimization approach (PR_GA_RF) for insulations’ thicknesses). The reduced brute-force search method

Fig. 15. Duration curves of the operative temperature (a) and the mean air temperature (b) at the warmest zone (upper floor) for the three studied cases.
122 M. Hamdy et al. / Building and Environment 46 (2011) 109e123

Fig. 16. Pareto fronts of the three studied cases (DH24  6250, 2400, and 1000  C h) and contribution of the heating/cooling system types in each case.

needs only 5  5  5  3  3  1 1 1 ¼1125 simulation runs. A of CO2-eq emissions and investment cost, considering the three
similar procedure for reduction was not applicable for the first two different levels of overheating in presenting the significant influence
studied Cases 1 and 2. of heating system type on the optimal solutions.
For Case 3, the obtained brute-force (1125 candidate solutions) From some observations in Fig. 16 at the value of CO2-eq emis-
is presented in Fig. 14 in square symbols while all the candidate sions (50 kg CO2-eq/m2 a), a higher investment cost is needed in
solutions which satisfy the thermal comfort limit (DH24  Case 2 compared with Case 1 in order to lower the overheating
1000  C h) are symbolized by circles. By indicating the obtained index DH24. This can be shown by points (A) and (B). Point (A) has
solutions of Case 3 in the figure, it can be seen that the obtained DH24 of 5400  C h which is higher than that for point (B), which is
solutions lie on the front side of the circle points (feasible solu- 2200  C h. For the latter, a lower building tightness, shading option
tions). This verifies the dominating nature of the obtained optimal and relatively thin insulation were implemented to decrease the
solutions. amount of summer overheating. However, this required a lot of
space heating energy (points A and B have space heating energies
8. Comparison of the three cases of 70 and 130 kWh/m2 a, respectively). To keep the same level of
CO2-eq emissions (equal environmental impact), the optimization
The preceding cases addressed three different levels of over- solver selected the Oil Fire Boiling system (Sys. 2) which has
heating (DH24  6250, 2400, and 1000  C h). Fig. 15 shows the a lower emission factor (EF ¼ 0.267 kg/kWh) for point (B) instead
duration curve of the operative and mean air temperatures at the of the Electrical Radiator (Sys. 1) (EF ¼ 0.459 kg/kWh) which was
warmest zone. The optimal solutions 41 (with the minimum the selection of point (A). This required an additional investment
overheating in Case 1, DH24 ¼ 4181  C h), 46 (with the upper bound cost of 9150 V (price difference between the heating systems 1
overheating in Case 2, DH24 ¼ 2400  C h), and 20 (with the upper and 2). Furthermore, the shading option costs point (B) additional
bound overheating in Case 3, DH24 ¼ 1000  C h) are selected, 4000 V to decrease the direct solar radiation. However, a 3150 V
respectively, to evaluate the temperature duration curves for the investment cost was saved by using less insulation and less effi-
three studied cases. cient building tightness. As a result, an additional cost (10 000 V)
The energy consumption of building depends significantly on the was needed for higher thermal comfort conditions (point B) to
criteria set for the indoor environment (e.g., thermal comfort crite- maintain the same impact on the environment (50 kg CO2-eq/
rion). However, the emission production is not only dependent on m2 a).
energy consumption but also on the heating/cooling energy source Finally, it is worthwhile to mention that, on average Case 1 used
as well as on the efficiency of the HVAC systems. Fig.16 compares the lower Ubldg (average Ubldg ¼ 0.24 W/m2 K) to attain minimum
optimal solutions (trade-off relations) of the studied cases in terms amounts of heating energy (consequently CO2-eq emissions).
 M. Hamdy et al. / Building and Environment 46 (2011) 109e123 123

However, Case 2 implemented higher Ubldg (average Acknowledgement

Ubldg ¼ 0.38 W/m2 K) for fewer summer overheating solutions
(DH24  2400  C h) with and without a cooling option. The inter- The authors would like to acknowledge the financial support of
mediate values of Ubldg (average Ubldg ¼ 0.3 W/m2 K) with the the Finnish National Technology Agency (TEKES), as part of the
cooling option are selected in Case 3. MASI programme, as well as the following supporting companies:
Optiplan Oy, Pöyry Building Servicees Oy, Saint-Gobain Isover Oy,
9. Conclusions Skanska Oy, and YIT Oyj.

Many of the building energy regulations have only building

envelope (U-values) and heat recovery requirements. The impor- References
tance of this study is that it extends the scope to include the supply
heating energy source and the shading option as well, aiming to [1] Vine E. Opportunities for promoting energy efficiency in buildings as an air
quality compliance approach. Energy 2003;28(4):319e41.
achieve low-emission cost-effective thermal-comfort solutions for [2] Carsten P, Boermans T, Harnisch J. Mitigation of CO2 emissions from the EU-15
modern buildings. building stock beyond the EU directive on the energy performance of build-
For this purpose, a modified multi-objective optimization ings. Ecofys GmbH. Eupener Straße, 59 50933 Cologne, Germany; 2006.
[3] Balaras Constantinos A, Droutsa Kalliopi, Dascalaki Elena,
approach is proposed and combined with IDA ICE (building
Kontoyiannidis Simon. Heating energy consumption and resulting environ-
performance simulation program) addressing three studied cases mental impact of European apartment buildings. Energy and Buildings
with different thermal overheating levels. A set of optimal solutions 2005;37:429e42.
[4] Balaras CA, Droutsa K, Argiriou AA, Asimakopoulos DN. Potential for energy
is obtained and analyzed. The analysis shows the physical meaning
conservation in apartment buildings. Energy and Buildings 2000;31
behind the optimizer’s selections (decisions) and helps to under- (2):143e54.
stand the simultaneous influence of the design parameters on the [5] Thyholt M, Hestnes AG. Heat supply to low-energy buildings in district
building emissions, investment cost, and thermal comfort. The heating areas: analyses of CO2 emissions and electricity supply security.
Energy and Buildings 2008;40:131e9.
most important conclusions are [6] Gustavsson L, Joelsson A. Life cycle primary energy analysis of residential
buildings. Energy and Buildings 2010;42:210e20.
 The influence of the supply heating system type on the results [7] Rolfsman B. CO2 emission consequences of energy measures in buildings.
Building and Environment 2002;37:1421e30.
is more significant than the influence of the other design [8] Hasan A, Vuolle M, Siren K. Minimisation of life cycle cost of a detached house
variables. Using a low-emission supply heating system reduces using combined simulation and optimization. Building and Environment
not only the emissions related to space heating energy, but also 2008;43(12):2022e34.
[9] Wang W, Zmeureanu R, Rivard H. Applying multi-objective genetic algorithms
the emissions related to domestic hot water and system heat- in green building design optimization. Building and Environment
ing loss energies, which are considerable in cold climate 2005;40:1512e25.
countries. [10] Verbeeck G, Hens H. Life cycle optimization of extremely low energy dwell-
ings. Journal of Building Physics 2007;31:143.
 For environmental and indoor thermal-comfort considerations, [11] Lee WL, Yik FWH. Regulatory and voluntary approaches for enhancing
investing in high-price low-emission supply heating system building energy efficiency. Progress in Energy and Combustion Science
could be a better solution than investing in additional insu- 2004;30(5):477e99.
[12] Hamdy M, Hasan A, Siren K. Combination of optimization algorithms for
lation and other low-energy measures.
a multi-objective building design problem. IBPSA: 11th International Building
 Additional insulation reduces the space heating energy Performance Simulation Association Conference, Glasgow-UK; 2009.
consumption. However, this increases the overheating during [13] The National Building Code of Finland C3. Thermal insulation in a building,
the summer. High efficient heat recovery units are a good regulations. Helsinki: Ministry of the Environment; 2007: p. 9.
[14] The National Building Code of Finland D5. Calculation of power and energy
solution for ‘low emission’, ‘low summer overheating’ dwell- needs for heating of buildings, guidelines. Helsinki: Ministry of the Environ-
ings. For economical and practical considerations, heat ment; 2007: p. 72.
recovery unit with annual efficiency in between 60 and 80% is [15] D2 Finland Code of building Regulation. Indoor climate and ventilation of
buildings, regulation and guidelines. Helsinki: Ministry of Environment;
a reasonable selection. 2003.
 Without considering the summer overheating level, the opti- [16] Pylsy P, Kalema T. Concepts for Low-Energy Single-Family Houses. Tampere
mization approach achieved a set of solutions which are University of Technology. Department of Mechanics and Design. Research
Report 1; 2008.
significantly better than the initial designs using additional [17] Heljo J, Laine H. Report 2. Institute of Construction Economics. Tampere
insulation. This means that additional insulation could be University of Technology: Finland; 2005.
required if operable windows (free cost solution) are available [18] www.tamar.fi.
[19] Deb K. Multi-objective optimization using evolutionary algorithms. John
for sufficient natural ventilation and cooling. Wiley and Sons; 2001.
 In cold climates, acceptable summer overheating levels could be [20] Sahlin P, Eriksson L, Grozman P, Johnsson H, Shapovalov A, Vuolle M. Whole-
achieved without a cooling option. However, cooling system building simulation with symbolic DAE equations and general purpose
solvers. Building and Environment 2004;39(8):949e58.
and shading option are required for high thermal comfort levels.
[21] Achermann M, Zweifel G. RADTEST radiant cooling and heating test cases.
A Report of Task 22, Subtask C. Building Energy Analysis Tools. Comparative
The simulation-based optimization approaches show a great Evaluation Tests, IEAdInternational Energy Agency, Solar Heating and Cool-
potential for the solution of multi-objective building design prob- ing Programme; April 2003.
[22] Crawley DB, Hand JW, Kummert M, Griffith BT. Contrasting the capabilities of
lems, and can be used in the design phase to give a better under- building energy performance simulation programs. In: Proceedings of the 9th
standing for the performance of the building and its HVAC systems. IBPSA Conference, Montreal, Canada; 2005. p. 231e238.