Energy and Buildings 102 (2015) 117

Contents lists available at ScienceDirect

Energy and Buildings

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

Innovative control logics for a rational utilization of electric loads and

air-conditioning systems in a residential building
G. Graditi a , M.G. Ippolito b , R. Lamedica c , A. Piccolo d , A. Ruvio c , E. Santini c , P. Siano d ,
G. Zizzo b,
a
ENEA Agenzia Nazionale per lEnergia, le Nuove Tecnologie e lo Sviluppo Economico Sostenibile, Research Center of Portici, P. E. Fermi, 1, 80055 Portici,
Naples, Italy
b
DEIM Department of Energy, Information Engineering and Mathematical Models, Universit di Palermo, viale delle Scienze, Palermo, Italy
c
Diaee Engineering Department of Astronautics, Electric and Energy, University of Rome Sapienza, Rome, Italy
d
DIIN Department of Industrial Engineering, Universit di Salerno, Via Giovanni Paolo II, 132, Fisciano (SA), 84084 Salerno, Italy

a r t i c l e

i n f o

Article history:
Received 5 December 2014
Received in revised form 13 May 2015
Accepted 18 May 2015
Available online 22 May 2015
Keywords:
Innovative control logics
Buildings controls
Energy efciency in buildings
Energy management systems
Smart house

a b s t r a c t
The paper focuses on the evaluation of the impact on residential buildings of innovative control logics
dened for a rational utilization of the electric loads and air-conditioning systems. The control logics are
implementable thanks to a Decision Support and Energy Management Systems developed by the same
authors and easily installable in new and still-existing houses.
The paper shows how a signicant reduction of the primary energy consumption can be obtained by
controlling the air-conditioning units together with the other electric loads of the house.
After having explained how the control logics operate, a simulation tool is described, developed for a
preliminary evaluation of their effects on the energy consumptions of residential buildings in the design
stage. An application of the simulation tool on a typical house, considering different scenarios and varying
the climatic conditions and the energy performance class of the envelope, is presented.
Finally, a measurement campaign on two real existing houses, equipped with the Decision Support and
Energy Management System, is described, and the effects of the control logics are assessed.
2015 Elsevier B.V. All rights reserved.

1. Introduction
In accordance with the European Standard EN 15232 [1], it is
well known how buildings energy performance can be inuenced
by the presence of Building Automation Control (BAC) Systems an
Technical Building Management (TBM) systems [24].
Also the Energy Performance of Building Directive (EPBD)
2010/31/EU [5] in line with Directive 2009/72/EC [6], promotes the
adoption of BAC and intelligent metering systems recognizing their
potentiality of reducing primary energy consumptions.
Moreover, in the last few years many research studies have
been devoted to Demand Side Management (DSM) policies [713]
implemented in order to achieve an intelligent and interactive
management of domestic devices and air-conditioning systems
obtaining, in this way, higher performances during normal operation. DSM is assessed as an integral part of the smart grid and one
of the most important methods of energy saving also thanks to new

Corresponding author. Tel.: +39 09123860205.

E-mail address: gaetano.zizzo@unipa.it (G. Zizzo).
http://dx.doi.org/10.1016/j.enbuild.2015.05.027
0378-7788/ 2015 Elsevier B.V. All rights reserved.

Information and Communication Technology (ICT) infrastructures

that allow faster and more efcient communications between users
and public utilities.
Recognizing the importance of this topic, the authors have
proposed some innovative control logics that, implemented in a
Decision Support and Energy Management System (DSEMS) for residential applications, can be applied to the operation of the electric
and thermal loads of a house for attaining the optimization of the
energy consumption.
Thanks to these control logics, the direct benets for customers
may be various, involving energetic, economic and comfort issues
(more exibility, different user proles according to the environment, elimination of overload black-out due to the simultaneous
use of more devices).
The utilization of the DESMS leads also to benets for the
distribution system operator (DSO): power losses reduction, possibility to defer the strengthening interventions of the existing
networks and to improve the employment of production units,
reduction of Green House Gases (GHG) emissions, etc. Moreover,
the load control can become a precious reserve in emergency
situations.

G. Graditi et al. / Energy and Buildings 102 (2015) 117

The control logics have been implemented also in the simulation tool TiDomus developed by the authors in Visual Basic (VB)
environment, carried out for assessing their effects while considering buildings with different characteristics and equipment. Then,
the same control logics have been implemented in two prototypes
of DSEMS installed in two real houses in order to carry out experimental tests.
The present study is part of National Research Project SIRRCE
(System for the resIdential eneRgy optimization with summeR air
Conditioning intEgration) supported by the Italian Ministry for the
Economic Development.
The aim of this paper is to illustrate the main results of the SIRRCE project, presenting the control logics, the simulation tool, the
DSEMS and the measurement campaign done on two real houses
in order to check the real operation of the DSEMS.
The structure of the paper is as follows. In Sections 2 and 3 the
control logics and the functioning of the real DSEMS are, respectively, described.
In Sections 4 and 5 TiDomus is, respectively, presented and
applied to a test case in different scenarios. Finally, in Section 6 the
results of a measurement campaign on two real houses are shown
while in Section 7 conclusions are given.
2. Innovative control logics
Many authors have proposed logics and strategies for the control and the management of electric and thermal loads in residential
and not residential buildings. In [14] the authors propose electrical loads management in emergency conditions, when main power
sources are limited or totally absent, for residential and commercial
buildings.
In [1517] four rule-based control logics and an articial neural
network (ANN)-based control logic are developed for the integrated
control of openings and cooling systems in buildings with double skin envelopes during summer. In [18] a fuzzy logic control
system to keep the illumination level at 350 lux in a room is presented. In [19] a lighting control algorithm is developed with the
goal of achieving energy efciency and health aspects of occupants
in ofce buildings in Dubai. In [20] an energy management strategy
based on the rules of the electricity bill for a commercial building
in a supermarket application is presented. Also in [20] the authors
establish some objectives as load shedding in order to reduce the
electricity bill and the CO2 emissions of commercial building, using
photovoltaic (PV) and storage systems.
In [21] the authors analyze the inuence of DSM strategies on
the performance of a thermally activated building system (TABS)
applied in a commercial building with the goal of estimating the
potential of TABS for load shifting requested by the electricity grid.
Finally, in [22] two different hydraulic congurations of heat
pump with thermal energy storage and four different control strategies are analyzed.
Most of the above-mentioned papers deal with specic loads
and, in particular, no one deals with the problem of the control of
single air-conditioning units for residential buildings.
The proposed control logics have been designed in order to face
this issue and in the same time to provide management strategies
for all the electric and thermal loads usually present in a house. The
control logics are implemented in a DSEMS for residential applications and the user may select and change one or more control
logic in every moment according to its preferences. Selecting a
given control logic among the six proposed, the end-user can attain
well-dened effects on its electrical and thermal load.
The proposed logics are:
Comfort;
Economy;

Fig. 1. Input and output of the DSEMS.

Emergency;
Energy;
Power;
Thermal storage.

The DSEMS is also able to receive control signals from the DSO by
activating a NET-Service (NetS) function. Therefore, NetS function
is not a proper control logic and is only selected when the end-user
wants the DSEMS to receive external signals. By activating the NetS
function, some selected electric and thermal loads can be controlled
by the DSO in order to achieve precise benets for the grid, while
the end-user will receive a premium or a discount of the electricity
price for the service it offers to the DSO.
The DSEMS, whose inputs and outputs are represented in Fig. 1,
has been described in detail in [23,24], presenting also the utilized
designing and testing methods.
The state machine that manages the activation of the control
logics consists of seven super-states. Each super-state handles a
different control logic (and the NetS function) and contains precise
sub-states and functions.
Even if the SIRRCE project was born with the aim of concentrating on summer air-conditioning, the DSEMS implements the
following actions in both summer and winter seasons:
indoor climate control during the economic tariff range: when
the power consumption exceeds the contractual power limit, the
temperature set point of the air-conditioning system is gradually
changed in order to reduce the energy absorption;
climate control during the expensive tariff range: the DSEMS
varies the temperature set point of the split in order to reduce
the energy absorption.
In the following it is described how each control logic operates.
2.1. Comfort
The Comfort function is selected in order to assure:
the energy supply to all types of loads;
the preservation of the maximum comfort in the house in terms
both of indoor temperature and of electrical load usage.
Controllable and shiftable loads are managed by the DSEMS in
order to avoid they exceed the maximum available active power,
limited by contract, thus improving the continuity of supply for the
end-user.

G. Graditi et al. / Energy and Buildings 102 (2015) 117

2.2. Economy
The Economy function is selected in order to achieve the minimum energy purchase cost. Every 15 min, the DSEMS automatically
determines the optimal electrical and thermal load conguration
able to guarantee the minimum cost and, simultaneously, the
respect of the system constraints.
2.3. Emergency
The Emergency function is automatically selected by the DSEMS
following a failure in the distribution grid supplying the user system.
In this case, the electrical supply is provided by a local generator
(photovoltaic, micro-wind, micro-turbine, etc.) and by an electric
energy storage system, suitably sized.
The DSEMS gives to the user the possibility of selecting manually
the Emergency scenario and to feed only the privileged loads of
the house disconnecting the user electrical system from the grid
(stand-alone mode).
2.4. Energy
The Energy function is selected in order to respect an assigned
consumption in terms of electrical energy or of monetary expenses
in a prexed period of time.The user selects the preferred consumption and the time period (a day, a week, a month, etc.).The DSEMS
sends messages to the user informing it about:
the daily-average consumption;
the possible consumption that are allowed in order to achieve the
prexed goal.
2.5. Power
The Power function is selected in order to guarantee that the
absorbed active power does not exceed a prexed threshold value.
The aims of this scenario are fundamentally two:
the user can assure to the DSO that its consumptions will never
exceed a xed maximum value in an assigned daytime period,
receiving from the DSO an economic benet;
the user can attain a reduction of its own consumptions.
2.6. Thermal storage
The Thermal storage function acts on the electric power
absorbed by the air-conditioning units, in relation to the change of
the preset temperature. This feature allows an anticipated cooling
in each controlled zone. The use of thermal storage can be considered as an alternative to electric storage; indeed it uses the inertia
of the environment to shift consumption for cooling. It is possible
to exploit the locally produced electrical energy and to reduce daily
load variations.
3. Implementation of the DSEMS
The DSEMS ensures the power supply of the loads, in respect
of the available power and performs detachments or control of the
electrical loads, when the available power is exceeded. In particular, it allows the control on specic loads selected by the end-user
following precise priorities. Based on these priorities, the system
performs sequences of detachments or control.
The control acts on all the electric load simultaneously (General) or on a specic load (Peculiar).

Some parameters dening the type of load operation are: normal, not interruptible cycle, thermal heat pump with split, with
thermal zone temperature control, etc.
The main adjustable or measurable system parameters are:

temperature: measured by the sensors within the temperature

control zone. In the presence of multiple probes, the reference is
the average temperature;
electric power absorbed by the loads: value detected by a dedicated meter located upstream of all loads;
locally generated electric power: value detected by a dedicated
meter placed downstream of the local generator;
electric power absorbed by the system: value obtained as the difference between the power absorbed by the loads and the locally
generated power. This parameter can assume negative values
when production exceeds consumption;
available electric power: value corresponding to the threshold
where is activated the loads detachment. This threshold can be
set during system conguration;
restoration electric power: the threshold value that allows the
reactivation of the loads;
activation thermal storage threshold: value below which the system activates the thermal storage;
detachment threshold of thermal storage: value above which the
system disables the thermal storage;
enable control: an input signal that allows the user enabling the
execution of the functions of automatic control.

The DSEMS performs a check cycle through switching control

and updates the status of the loads with uninterruptible cycles. If
the DSEMS veries an energy request below the threshold, in the
case of detached loads, it restores the loads one at a time starting from the one with the highest priority. On the contrary, if the
energy request is above the threshold, the DSEMS enables the load
detachment function. The load detachment order is from the lowest
priority load to the highest priority one, with three levels.
In the rst level the DSEMS detaches the ordinary loads with
interruptible working cycle, with working cycle started on a time
less than a select threshold (e.g. 15 min), or the load whose working
cycle is nished.
In the second level the DSEMS detaches the thermal loads and
the loads with a not interruptible working cycle in progress.
In the third level the loads detachment works subjected to forcing, set manually by the user who has forced the switching of a
load, as a result of a detach, or has forced the load to avoid a detachment of level with priority less than three and loads that had not
been turned off in the previous level.
If the Comfort function is activated, only a power threshold
check is performed by the DSEMS.
If the user activates the Thermal storage function, the system
performs an additional check: if the energy production from a local
source exceeds the house consumption, the temperature set point
of the air-conditioning systems is changed and the air conditioner
is switched on.
By activating the Economy function, the system allows load
detachments only in high electricity cost time slots.
With the Energy function, the system allows displaying the
average consumption from the beginning of the month (kWh or
euro); this average is compared with the threshold xed by the
end-user. If the threshold is set in kWh, the system calculates the
hourly available rate from the energy threshold set by the user.
This hourly available rate is calculated for each hour by the amount
of energy that is available from the beginning of the month. The
information is provided to the user with light signals:

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 2. Interface for enabling functions and for entering the thresholds of the parameters for the Power and Energy functions.

green, if the consumption is less than 95% of the available energy;

yellow, if the consumption is in the range 95105% of the available energy;
red, if the consumption is higher than 105% of the available
energy.
If the user chooses to keep costs under control, the steps
reported above are evaluated on the basis of the energy cost based
on the users contract type, with different tariff in the day.
Enabling the Power function, a power limit chosen by the user,
lower than the contractual threshold is adopted.
Finally, enabling the Emergency function, in the case of grid
failure, if a local generator and an electric storage are available, the
system will send an opening command to a normally closed contactor installed in the apartment distribution board, detaching the
ordinary loads and supplying only the privileged loads. Following
the grid recovery, the DSEMS sends a closing signal to the contactor
restoring the supply also of the ordinary loads.
Fig. 2 illustrates the webpage available at the operation center,
from which the end-user can enable or disable individual features
and enter the threshold for the Power and Energy functions.
4. Calculation of the electric energy consumption
The control logics described in Section 2 have been implemented
in a simulation tool, named TiDomus, that is able to reproduce
different house environments by varying:
the type and the nominal power of the electric loads;
the thermal characteristics of the building envelope;
the type and the technical characteristics of the thermal/cooling
system;
the presence/absence of inhabitants in the house.
TiDomus can also take into consideration the presence of a local
generator or an electric energy storage system.
TiDomus has been developed in VB environment with the aim of
providing to the designer of the BAC system a tool for evaluating the
energy and economic savings due to the installation of the DSEMS,
implementing the above-described innovative control logics.

TiDomus uses a probabilistic approach for the construction

through a Monte Carlo extraction process of the daily power prole
of the house starting from the knowledge of some social, economic
and demographic factors [2532].
Being TiDomus based on SirSym-Home presented in [4], the
daily load prole of the house is found according to the bottomup approach dened in [25], as sum of all the power proles of the
electric loads.
For this reason the calculation is inuenced by some factors
widely variable from house to house, such as the number of
occupants, the variability of their habits with the seasons or the
possibility of choosing different working cycles for some domestic
appliances.
On the contrary, the daily thermal prole of the house is deterministically calculated, by solving the general equations of the
thermal exchange for the system and by applying the UNI TS 11300
Standards [33,34]. UNI TS 11300 Standards are based on EN ISO
13790 Standard [35] and are mandatory in Italy for buildings energy
needs calculations.
Starting from the daily power prole of the house, TiDomus calculates the yearly energy consumption dividing the year into three
periods: summer, winter and intermediate (autumn and spring).
The outputs of the simulation tool are:

the daily/yearly electric energy saving;

the daily/yearly thermal energy saving;
the yearly economic saving;
the yearly reduction of GHG emissions.

The yearly economic saving is calculated knowing the electricity

cost paid by the end-user, considering also its variability during the
day.
The architecture of the simulation tool, represented in Fig. 3,
presents:
a MAIN routine managing the interface with the user, the data
loading, the output and the interaction among the various subroutines;
the Electric Load Simulator (ELS) subroutine, building the daily
electric load prole;

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 3. Architecture of TiDomus.

Fig. 4. TiDomus: mask for the selection of the electric loads.

the Thermal Behavior Simulator (TBS) subroutine, building the

thermal electric load prole;
the Electric Source Simulator (ESS) subroutine, allowing the calculation of the electric energy produced by a photovoltaic system;
the Control Logic Simulator (CLS) subroutine, implementing the
innovative control logics.

the mask for the specication of the electricity price for the enduser in every hour of the day;
the mask containing the results of the calculations.

5. Case study
Differently from SirSym-Home presented in [4], realized only for
research purposes and having a not easily understandable graphic
interface, TiDomus has been realized for being used by power system designers and energy consultants.
In Figs. 47 some screenshots of TiDomus are shown, in particular:
the mask for the specication of the electrical loads (number and
nominal power);
the mask for the specication of the energy performance class of
the building according to the EN 15232 Standard [1];

The effect of the control logics has been rstly evaluated considering the 140 m2 test house reported in Fig. 8.
Having effects on the thermal calculations, three different locations characterized by three different latitudes have been chosen
for the house: Palermo in the South of Italy, Rome in the Middle
and Turin in the North. The climatic data for the three cities were
taken from the Italian Standard UNI 10349 [36] and are reported in
Table 1.
The electric loads present in the house are listed in Table 2.
Air-conditioner units in the various rooms are sized in order
to provide the total thermal/cooling load of each room calculated
according to UNI TS 11300 [33,34].

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 5. TiDomus: mask for the calculation of the primary energy requirements.

Fig. 6. TiDomus: mask for the evaluation of the economic saving and of the avoided CO2 emissions.
Table 1
Climatic data for simulations (UNI 10349).
City

Latitude

Reference summer external

temperature
( C)

Reference winter external

temperature
( C)

Palermo
Rome
Turin

38 06
41 54
45 04

32
33
30.5

5
0
8

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 7. TiDomus: mask containing the nal economic evaluation.

The effects of the DSEMS action on the electric energy consumption of the house are evaluated:
considering four of the six control logics previously dened and
the NetS function (Emergency and Thermal storage have not been
considered because they do not have the aim of achieving energy
savings);
considering that the DSO imposes a 2 kW upper limit to the power
absorbed by the system when the NetS function is active;
varying the energy performance class of the house from A to G;
considering only a winter day (WD) and a summer day (SD).
Simulations are done considering that the duration of the daily
activation time of the air-conditioner system is different during
winter period for the three cities, according to the Italian Decree

Fig. 8. Floor plan of the test house (case study 1).

DPR 412/93 [37], as reported in Table 3. On the contrary, the simulations concerning the summer period are done considering the
air-conditioner system active 8 h per day for every location.
In Table 4 the electric energy consumptions of the house in the
absence of the DSEMS action are reported (base scenario).
Varying the efciency class of the house the electric energy consumptions vary in the range 50007000 kWh/year, that is a realistic
range in the presence of domestic loads as the dryer and the electric
storage water heater.
In Tables 59 the electric energy consumptions of the house in
the presence of the control logics implemented in the DSEMS are
reported.
Table 2
List of the electric loads of the test house (case study 1).
Category

Description

Electric power
request during
normal operation
(W)

Stand-by
consumption
(W)

Lighting
system

Entrance lighting
Living room lighting
Kitchen lighting
Corridor lighting
1st Toilet lighting
2nd Toilet lighting
1st Bedroom lighting
2st Bedroom lighting
Outdoor lighting

60
54
36
120
100
100
100
100
36

0
0
0
0
0
0
0
0
0

Appliances

Polisher
Hair dryer
Electric shaver
Electric iron
Toaster
Electric oven
Microwave oven
Fridge-Freezer
Dishwasher
Washing machine
Balanced ventilation
system
Electric dryer
Exhaust fan

1000
1200
40
1200
1000
2000
2000
350
2700 (max)
2000 (max)
400

0
0
0
0
0
0
3
2
0
0
0

2700 (max)
150

0
0

Entertainment PC
HI-FI
system
TV, VCR and DVD
player
Doorphone
Electric storage water
Others
heating

120
300
400

5
9
21

40
1800

25
0

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Table 3
Duration of the daily activation time of the air-conditioning system during the heating period, according to DPR 412/93 Section 9 [37].
City

Climatic zone

Heating period

Duration of the
activation time
[h/day]

Palermo
Rome
Turin

B
D
E

From 1st of December

From 1st of November
From 15th of October

To 31st of March
To 15th of April
To 15th of April

8
12
14

Table 4
Daily energy consumption of the test house for different energy classes and locations of the house (base scenario).
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.86
15.90
16.41
16.90
17.40
17.89
18.38

15.42
15.91
16.87
17.84
18.81
19.81
20.72

14.88
15.30
15.53
16.29
17.01
17.50
17.52

16.98
17.68
18.75
20.69
22.22
24.25
25.18

15.65
15.93
16.21
16.50
16.80
17.10
17.37

19.98
21.68
23.75
25.69
27.22
29.25
31.18

Figs. 914 graphically synthetize the results of the calculations

in Tables 59. For each location, for each period of the year and for
every energy class, the daily energy consumption for each scenario
is represented as the percentage of the daily energy consumption
in the base scenario.

In every examined case, the daily energy consumption is equal

or lower than in the base scenario. Starting from the data in
Tables 49, the average percentage reductions of the daily energy
consumption with respect to the base scenario are reported in
Table 10.

Table 5
Daily energy consumption of the test house, in the comfort scenario, for different energy classes and locations of the house.
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.86
15.90
16.41
16.88
17.17
17.33
17.19

15.42
15.91
16.85
17.77
18.67
19.58
20.35

14.88
15.30
15.53
16.13
16.52
16.34
16.50

16.98
17.68
18.75
20.69
22.18
24.14
25.01

15.65
15.93
16.14
16.18
16.03
17.10
17.37

19.98
21.68
23.75
25.53
26.73
28.09
30.16

Table 6
Daily energy consumption of the test house, in the energy scenario, for different energy classes and locations of the house.
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.86
15.90
16.41
16.86
17.18
17.33
17.08

15.42
15.91
16.83
17.79
18.70
19.59
20.41

14.88
15.30
15.50
16.12
16.51
16.33
16.35

16.98
17.68
18.74
20.66
22.17
24.14
25.00

15.65
15.93
16.14
16.18
16.03
17.08
17.24

19.98
21.68
23.72
25.52
26.72
28.08
30.01

Table 7
Daily energy consumption of the test house, in the economy scenario, for different energy classes and locations of the house.
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.45
15.31
15.27
15.53
15.89
16.90
16.92

14.96
15.32
16.15
16.98
17.89
18.83
19.68

14.59
14.67
14.55
15.08
15.44
16.08
15.81

16.61
17.24
18.08
19.91
21.39
23.34
24.24

15.45
15.20
15.28
15.38
15.15
15.26
15.33

19.69
21.05
22.77
24.48
25.65
27.83
29.47

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Table 8
Daily energy consumption of the test house, in the power scenario, for different energy classes and locations of the house.
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.86
15.89
16.13
16.16
15.45
15.97
15.08

15.32
15.75
16.57
17.45
18.30
19.17
19.83

14.86
15.17
15.02
15.15
14.90
15.20
15.42

16.94
17.58
18.60
20.47
21.93
23.85
24.63

15.60
15.67
15.41
14.90
14.94
14.99
15.30

19.96
21.55
23.24
24.55
25.11
26.95
29.08

Table 9
Daily energy consumption of the test house, with nets function activated, for different energy classes and locations of the house.
Energy class

A
B
C
D
E
F
G

Palermo

Rome

Turin

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

SD
[kWh]

WD
[kWh]

14.43
15.32
15.67
15.97
16.38
16.77
17.17

14.96
15.34
16.18
17.03
17.96
18.92
19.79

13.48
13.66
13.63
14.06
14.50
14.72
14.56

15.79
16.43
17.42
19.23
20.69
22.62
23.52

13.20
13.08
12.96
12.86
12.76
12.66
12.54

18.58
20.04
21.85
23.46
24.71
26.47
28.22

Table 10
Average percentage variation of the energy consumption with respect to the base scenario.
Scenario

Comfort
Economy
Energy
Power
NetS

Palermo

Rome

Turin

SD

WD

SD

WD

SD

WD

2%
6%
2%
7%
5%

1%
4%
1%
2%
4%

2%
7%
3%
7%
13%

0%
3%
0%
1%
7%

1%
7%
1%
7%
22%

1%
1%
4%
4%
9%

Fig. 9. Percentage daily energy consumption (Palermo, Summer).

Fig. 10. Percentage daily energy consumption (Palermo, Winter).

By analyzing the data in Table 10, it can be easily seen that the
NetS function produces the highest reduction of the energy consumption while the Comfort scenario produces the lowest ones.
Nevertheless the NetS function is not an independent scenario
but rather a function activated by the DSO and able to overlie the
specic control logics of the other scenarios. Therefore, excluding
for this reason the results related to the NetS function, Economy is
the scenario that gives place to the highest reduction of the energy
consumption for the end-user.
Moreover, Tables 49 show that, in the presence of very low
energy class of the building envelope, the highest reductions in the
energy consumptions occur.

For example, in Fig. 9, a reduction of about 18% of the initial

energy consumption is obtained for class G whilst a reduction of
about 2% is obtained for class A in the Power scenario.
Therefore, a general conclusion can be drawn from the results
of the calculations: the installation of a DSEMS and the application
of specic control logics for energy consumption optimization are
more effective and convenient as higher it is the energy consumption of the building and as lower it is the initial energy class, as
highlighted also in [4].
Simulations show that climatic conditions inuence the effects
of the control logics. In all the simulations, outdoor temperatures
have been assumed constant during the whole day and equal to
those in Table 1.

10

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 11. Percentage daily energy consumption (Rome, Summer).

Fig. 13. Percentage daily energy consumption (Turin, Summer).

Fig. 12. Percentage daily energy consumption (Rome, Winter).

Fig. 14. Percentage daily energy consumption (Turin, Winter).

The assumption of constant temperature has been done for

showing more clearly the effect of the climatic conditions on
the control logics, without introducing temperature proles that
can have very different shapes from place to place. This working condition, even if not fully realistic, is usually accepted in

literature [36,37] for heating/cooling calculation done on the side of

safety. Anyway, TiDomus allows also the implementation of specic
hourly temperature proles.
During summer period, the highest percentage reductions of the
energy consumption are found for Rome, that is the city, among

Fig. 15. Cant site daily consumptions and production in July.

G. Graditi et al. / Energy and Buildings 102 (2015) 117

11

Fig. 16. Cant site daily consumptions and production in August.

the three considered ones, having the highest medium external

temperature.
During winter period, the highest percentage reductions of the
energy consumption are found for Turin, that is the city, among the
three considered ones, having the lowest medium external temperature and the highest duration of the activation time of the heating
system.

6. Measurement campaign on two residential sites

A measurement campaign has been performed on two residential sites with different types of electrical equipment, and different
contractual power, for the months of July, August and September

2013 in order to test and validate the algorithms and management

policies implemented by the DSEMS [38].
The residential sites are located in two small cities of the Northern Italy.
The rst site, located in Cant, near Comos lake, is a cottage
of 160 m2 (with energy performance class B). The installed electric
power capacity is 6 kW with a 6 kW rated power PV system and a
cooling system controlled by a thermal power station. The management policy carried out on the general system control is a Thermal
storage function.
The second site is an apartment with a surface of 95 m2 in Mariano Comense (with energy performance class D). The installed
electric power capacity is 3 kW with split air-conditioning control.
The Economy function for loads displacement during the night
time is enabled.

Fig. 17. Cant site daily consumptions and production in September.

12

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 18. Cant site with PV production. August 7, 2013.

6.1. Measures
The data acquisition, necessary to verify the validity of the
management policies, was conducted with a system of monitoring and tests, controlled by a control unit that stores data

and events related to the system in which it is installed.

These les can be remotely downloaded by connecting the
Internal network via the ADSL modem present in the installation. This system acquires, every 15 min, the mean values
of:

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 19. Mariano Comense site daily consumptions in July.

Fig. 20. Mariano Comense site daily consumptions in August.

13

14

G. Graditi et al. / Energy and Buildings 102 (2015) 117

Fig. 21. Mariano Comense site daily consumptions in September.

the indoor temperature for each temperature control zone;

the total power required by the installation;
the power required by each load;
the power required by the air-conditioning system in each temperature control zone.

A logging system provides the events registration related to the

operation of the functions on the system (activation, detachments
of electrical and thermal loads).
The data shown in the graphs illustrate:
the power required by the individual loads, produced by the local
source (if present) and the requirements of the entire system;
the external and internal temperature proles of individual
areas/zones and the temperature set-point managed by the control system;
the actions performed by the control system on individual users
or controlled areas.
6.2. Residential site with PV production 6 kW
The management policies enabled in the Cant apartment
are able to use the excess electrical energy generated by the
photovoltaic system for thermal storage by increasing power consumption of heat pumps for an anticipated cooling of the interested
zones.
The programmed temperature prole is set every day at 26 C,
from 24.00 p.m. to 06.00 a.m., at 30 C, from 06.00 a.m. to 18.00 p.m.,
and at 26 C, from 18.00 p.m. to 24.00 a.m. Detachment is set with
a temperature of 30 C.
The Thermal storage is enabled during the whole day. With a
change of 6 C, the control system modies the preset temperatures for each time slots. The thermal storage function works if the
difference between electrical energy consumption and PV electrical energy production is less than 1000 W and is disabled if this
difference exceeds 0 W.

The loads priorities set in the DSEMS are:

1.
2.
3.
4.
5.
6.
7.
8.
9.

Electric storage water heater;

Bathroom heater;
Zone 1 Bathroom;
Zone 3 Bedroom;
Zone 4 Living area;
Zone 5 Kitchen;
Washing machine;
Dishwasher;
Electric oven.

If the available power is exceeded, load detachment occurs hierarchically from the last priority to the rst one. The restoration
occurs from the rst priority to the last one and takes place when
the input power is below the reset threshold. Figs. 1517 show the
total energy consumption and PV production and the consumption
of each load in the whole measurement period.
From the data analysis it can be seen that the highest electricity consumption is due to the air-conditioner system that, in some
days, requires more than 20 kWh (e.g. 27, 28 and 29 July).
Exemplifying graphs (Fig. 18) report the most relevant measurement of the residential site with a PV production recorded on 2013,
7th August. In particular:
the rst pair shows the power proles of conditioning and general, representing the difference between total consumption and
the PV production (a1a2 of Fig. 18);
the second pair shows the power proles of PV production and
total consumption (b1b2 of Fig. 18);
the third pair shows the outdoor temperature and the temperature proles of the zone 4 living (temperature prole, measured
temperature, temperature set-point) (c1c2 of Fig. 18).
The energy provided by the PV system is always higher than
the total energy request and part of this energy is used for a

G. Graditi et al. / Energy and Buildings 102 (2015) 117

15

Fig. 22. Mariano Comense site. July 25, 2013.

Thermal storage. Fig. 18 a1 shows that the activation of the

Thermal storage function occurs in the time period from 10 a.m.
to 12 a.m., when the energy not consumed (General) exceeds the
threshold set in the system management (1000 W). The activation of the function is easily seen in the graph of the temperature
(Fig. 18 c2) in the zone 4 living, which sets forth the change
in the set-point (from 30 C to 24 C). Correspondingly, there is an
increase in consumption for the heat pump (Fig. 18 a1).
The difference between the PV production and the total consumption also appears to be always in excess, in spite of this, the
energy drawn from the grid is always present because the auto

energy consumed (energy from PV energy available to feed into

the grid) is lower than the total energy consumed. The share of
self-consumption is around 4060% with respect to the total consumption.
6.3. Residential site 3 kW
The management policies enabled in residential sites aim at
containing the overloads of the system due to the contractual
power of 3 kW. In particular, for the Mariano Comense apartment,
Economy function is enabled, and the dishwasher and washing

16

G. Graditi et al. / Energy and Buildings 102 (2015) 117

machine are activated in the night (from Monday to Friday after

20:00 p.m. until 7:00 a.m.).
The load priorities set in DSEMS are:
1.
2.
3.
4.
5.
6.
7.
8.

Study split;
Dryer;
Bedroom split;
Bedroom split;
Split stay;
Washing;
Dishwasher;
Oven.

The thermal prole is enabled from Monday to Friday: 30 C

from 0:00 a.m. to 18:00 p.m., 25 C from 18.00 p.m. to 24.00 a.m.,
Saturday to Sunday: 30 C from 0.00 a.m. to 10.00 a.m., 25 C from
10.00 a.m. to 24.00 a.m. Detachment is set with a temperature of
30 C.
Figs. 1921 show the total energy consumption and the consumption of each load in the whole measurement period.
Exemplifying graphs in Fig. 22 report the most relevant measurement recorded on 2013, 25th July. In particular:
the rst pair shows the power prole of the four split and the
action of the control system to manage four split consumption
(a1a2 of Fig. 22);
the second pair shows the power proles general and dishwasher (b1b2 of Fig. 22);
The graphs (Fig. 22 a1a2) show the consumption (a1) and
the management (a2) of the air conditioning units. The goal of the
DSEMS is to limit the power required by the splits in order to avoid
that the total power required by the apartment loads exceeds the
contractual power threshold (PMax = 3 kW in graph b1). Graph
b2 of Fig. 22 shows how the Economy function enables the
dishwasher during night time (switch-on time after 21:00).
7. Conclusion
DSM, involving both electric and thermal loads, is a very interesting issue for researchers and industry [3942]. For this reason,
some innovative control logics for electric and thermal loads,
implementable in residential buildings thanks to a DSEMS have
been presented.
The control logics proposed in the present paper can operate in
real-time and do not need predictive models for assuring the energy
saving objectives. The absence of predictive models implemented
in the DSEMS, makes the designed system a very exible tool, easily
programmable for including new rules or upgrading the control
logics.
In the paper, the control logics have been tested using a simulation tool, named TiDomus, carried out by the same authors and
able to assess their effects on residential buildings having various
characteristics and equipment.
TiDomus has been used for calculating the reduction of the
energy consumption of a test house varying the energy class, the
location and the period of the year.
Moreover, two of the control logics implemented in the DSEMS,
have been tested on a real existing house, and the measurement
campaign done in the period from June to September 2013 has been
presented, showing the operation of the DSEMS. Being the focus of
the SIRRCE research project on summer air-conditioning, the measurement campaign concerned only Italian summer period and no
eld measurements have been done during winter season.

Simulations have shown that the installation of a DSEMS can

reduce the electricity bill but also the thermal energy consumptions
of the house and that this effect is as more signicant as lower is
the energy performance class of the building, and as higher is the
duration of the activation time of the air-conditioning system. This
result is perfectly in line with the study presented in [4] on BAC and
TBM systems for domestic installations.
Also, the effects of the outdoor temperature has been examined. In particular, it has been shown that energy savings are the
highest in the locations characterized by the most extreme outdoor
temperatures (e.g. Rome during summer and Turin during winter
period).
This is a very important result because it allows afrming that
simplied energy savings evaluation methodologies as the BAC factor method proposed by EN standard 15232 [1], can be improved
considering external climate conditions and duration of the activation time of the heating/cooling system.
Moreover, although some of the proposed control logics give
place to energy savings lower than those obtainable with others, a
correct evaluation of their advantages for the end-user should take
into account also the different limitations in the utilization of the
house electric and thermal loads. For example the Comfort control logic gives the lowest energy savings but assures no limitations
in the use of the loads.
While obtaining the possible benets of BAC, having high
costs, operation scenarios have utmost importance and hence, they
should be applied, properly.
Anyway, it is important to underline that the NetS scenario can
be implemented in various ways, given that it has been introduced
in order to allow the DSEMS receiving command signals from the
DSO. Indeed, the implementation of the ancillary service market
will promote new Demand Response (DR) policies that could be
easily implementable using simple and exible devices, like the
DSEMS presented in the paper operating in NetS modality.
Finally, it is reasonable expecting that the potentiality of the
proposed system could be increased by the aggregation of a great
number of users equipped with it.
For example, the FP7 ADDRESS project [43], the Italian National
i-NEXT project [44] and the CECED contract [45], have analyzed
also the effects of DSM and DR policies on the electricity market
and on the efciency of the whole electrical system, highlighting
how important can be the implementation of specic DR control
logics involving clusters of end-users and producers in the same
energy district.

References
[1] European Technical Standard EN 15232, Energy Performance of Buildings
Impact of Building Automation, Control, and Building Management, second
ed., CEN, Brussels, 2012.
[2] A. Dounis, P. Tiropanis, A. Argiriou, A. Diamatis, Intelligent control system for
reconciliation of the energy savings with comfort in buildings using soft
computing technique, Energy Build. 43 (1) (2011) 6674.
[3] J. Figueiredo, J. S da Costa, A SCADA system for energy management in
intelligent buildings, Energy Build. 49 (2012) 8598.
[4] M.G. Ippolito, E. Riva Sanseverino, G. Zizzo, Impact of building automation
control systems and technical building management systems on the energy
performance class of residential buildings: an Italian case study, Energy Build.
69 (2014) 3340.
[5] Directive 2010/31/EU of the European Parliament and of the Council of 19
May 2019 concerning the energy performance of buildings.
[6] Directive 2009/72/EC of the European Parliament and of the Council of 13 July
2009 concerning common rules for the internal market in electricity.
[7] M.J. Koer, C. Reinisch, W. Kastner, A semantic representation of
energy-related information in future smart homes, Energy Build. 47 (2012)
169179.
[8] A. Prudenzi, A. Silvestri, R. Lamedica, M.C. Falvo, U. Grasselli, M. Regoli,
Standby power end users and their demand management impact on
households load shape, in: IEEE Power and Energy Society General Meeting,
2429 July, 2011, pp. 16.

G. Graditi et al. / Energy and Buildings 102 (2015) 117

[9] P. Siano, Demand response and smart grids a survey, Renew. Sustain. Energy
Rev. 30 (2014) 461478.
[10] A. Gomes, A. Gomes Martins, R. Figueiredo, Simulation-based assessment of
electric load management programs, Int. J. Energy Res. 23 (1999) 169181.
[11] A. Zakariazadeh, S. Jadid, P. Siano, Economic-environmental energy and
reserve scheduling of smart distribution system: a multiobjective
mathematical programming approach, Energy Convers. Manage. 78 (2014)
151164.
[12] A. Zakariazadeh, H. Omid, J. Shahram, P. Siano, A new approach for real time
voltage control using demand response in an automated distribution system,
Appl. Energy 117 (2014) 157166.
[13] A. Zakariazadeh, S. Jadid, P. Siano, Multi-objective scheduling of electric
vehicles in smart distribution system, Energy Convers. Manage. 79 (2014)
4353.
[14] R. Lamedica, E. Santini, S. Teodori, D.Z. Romito, Electrical loads management
in energy emergency conditions, Electr. Power Energy Syst. 66 (2015) 8696.
[15] J.W. Moon, J. Leeb, Y. Yoon, S. Kim, Determining optimum control of double
skin envelope for indoor thermal environment based on articial neural
network, Energy Build. 69 (2014) 175183.
[16] J.W. Moon, J. Leeb, J.D. Chang, S. Kim, Preliminary performance tests on
articial neural network models for opening strategies of double skin
envelopes in winter, Energy Build. 75 (2014) 301311.
[17] J.W. Moon, J. Leeb, S. Kim, Application of control logic for optimum indoor
thermal environment in buildings with double skin envelope systems, Energy
Build. 85 (2014) 5971.
[18] S. Grgl, Energy saving in lighting system with fuzzy logic controller which
uses light-pipe and dimmable ballast, Energy Build. 61 (2013) 172176.
[19] P.K. Soori, M. Vishwas, Lighting control strategy for energy efcient ofce
lighting system design, Energy Build. 66 (2013) 329337.
[20] H. Zhang, A. Davigny, F. Colas, Y. Poste, B. Robyns, Fuzzy logic based energy
management strategy for commercial buildings integrating photovoltaic and
storage systems, Energy Build. 54 (2012) 196206.
[21] A. Arteconi, D. Costola, P. Hoes, J.L.M. Hensen, Analysis of control strategies for
thermally activated building systems under demand side management
mechanisms, Energy Build. 80 (2014) 384393.
[22] U.I. Dara, I. Sartorib, L. Georgesa, V. Novakovica, Advanced control of heat
pumps for improved exibility of Net-ZEB towards the grid, Energy Build. 69
(2014) 7484.
[23] G. Graditi, M. Atrigna, A. Piccolo, P. Siano, Energy management system for
smart homes: testing methodology and test-case generation, in: ICCEP 2013,
Alghero, Italy, 1113 June, 2013, pp. 766771.
[24] P. Siano, G. Graditi, M. Atrigna, A. Piccolo, Designing and testing decision
support and energy management systems for smart homes, J. Ambient Intell.
Humaniz. Comput. 4 (6) (2013) 651661.
[25] A. Prudenzi, A. Silvestri, R. Lamedica, M.C. Falvo, M. Regoli, Residential DSM
actions impact prediction through a psychological model of electricity, in: Int.
Conference IEEE Power and Energy Society General Meeting, 2226 July,
2012, pp. 16.
[26] A. Mahdavi, F. Tahmasebi, Predicting peoples presence in buildings: an
empirically based model performance analysis, Energy Build. 86 (2015)
349355.
[27] A. Campoccia, E. Riva Sanseverino, G. Zizzo, A Monte Carlo approach for a
study on the impact of the domestic installation of small PV and thermal solar

[28]
[29]
[30]
[31]

[32]

[33]
[34]

[35]

[36]
[37]

[38]

[39]

[40]
[41]

[42]

[43]

[44]
[45]

17

systems on the grid, in: 7th World Energy System Conference, WESC,
Romania, 2008, pp. 16.
T.A. Nguyen, M. Aiello, Energy intelligent buildings based on user activity: a
survey, Energy Build. 56 (2013) 244257.
B. Poel, G. van Cruchten, C.A. Balaras, Energy performance assessment of
existing dwellings, Energy Build. 39 (4) (2007) 393403.
K. Basu, L. Hawarah, N. Arghira, H. Joumaa, S. Ploix, A prediction system for
home appliance usage, Energy Build. 67 (2013) 668679.
S. Karjalainen, Should it be automatic or manual the occupants perspective
on the design of domestic control systems, Energy Build. 65 (2013)
119126.
G. Wood, M. Newborough, Energy-use information transfer for intelligent
homes: enabling energy conservation with central and local displays, Energy
Build. 39 (2007) 495503.
U NI TS 11300-1, Energy Performance of Buildings Part 1 Calculation of
Energy Use for Space Heating and Cooling, second ed., UNI, Milano, 2014.
U NI TS 11300-2, Energy Performance of Buildings Part 2 Calculation of
Energy Primary and Energy Performance for Heating Plant and Domestic Hot
Water Production, second ed., UNI, Milano, 2014.
European Technical Standard EN ISO 13790, Energy Performance of Buildings
Calculation of Energy Use for Space Heating and Cooling, rst ed., CEN,
Brussels, 2008.
UNI 10349, Riscaldamento e raffrescamento degli edici. Dati climatici (in
Italian), rst ed., UNI, Milano, 1994.
D.P.R. n. 412, Regolamento recante norme per la progettazione, linstallazione,
lesercizio e la manutenzione degli impianti termici degli edici ai ni del
contenimento dei consumi di energia, in attuazione dellart. 4, comma 4, della
L. 9 gennaio 1991, n. 10 (in Italian), 1993, August.
U. Grasselli, R. Lamedica, A. Ruvio, D. Novati, Management policies for
residential and tertiary users: a measurement campaign, in: 2014
International Symposium on Power Electronics, Electrical Drives, Automation
and Motion, SPEEDAM 2014, Ischia, Italy, 1820 June, 2014.
A. Arteconia, D. Costolab, P. Hoesb, J.L.M. Hensen, Analysis of control
strategies for thermally activated building systems under demand side
management mechanisms, Energy Build. 80 (2014) 384393.
J.K. Grubera, M. Prodanovica, Two-stage optimization for building energy
management, Energy Proc. 62 (2014) 346354.
M. Puchegger, Electric load behaviour and DSM potential of ofce buildings,
Energy Build. (2015), http://dx.doi.org/10.1016/j.enbuild.2014.12.046 (in
press).
F. Brahman, M. Honarmand, S. Jadid, Optimal electrical and thermal energy
management of a residential energy hub, integrating demand response and
energy storage system, Energy Build. 90 (2015) 6575.
ADDRESS: Active Distribution network with full integration of Demand and
distributed energy RESourceS, European Communitys Seventh Framework
Programme (FP7/2007-2013) grant agreement n 207643, http://www.
addressfp7.org/index.html
i-NEXT: Innovation for greeN Energy and eXchange in Transportation, Italian
project PON04a2 H.
G. Ala, V. Cosentino, A. Di Stefano, G. Fiscelli, F. Genduso, G.C. Giaconia, M.G.
Ippolito, D. La Cascia, F. Massaro, R. Miceli, P. Romano, C. Spataro, F. Viola, G.
Zizzo, Energy Management via Connected Household Appliances, rst ed.,
McGraw-Hill, Milano, 2008.