Upload
jun005
View
217
Download
0
Embed Size (px)
Citation preview
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
1/13
Modeling and optimization of a hybrid system for the energy supply
of a Green building
Hanane Dagdougui a,b, Riccardo Minciardi a, Ahmed Ouammi c,, Michela Robba a, Roberto Sacile a
a Department of Communication, Computer and System Sciences (DIST), Faculty of Engineering, University of Genoa, Genoa, Italyb MINES ParisTech, Sophia Antipolis Cedex, Francec Unit des Technologies et Economie des Energies Renouvelables, CNRST BP 8027 NU Rabat, Morocco
a r t i c l e i n f o
Article history:
Received 28 July 2011
Received in revised form 23 May 2012
Accepted 27 May 2012
Available online 27 September 2012
Keywords:
Decision support systems
Optimization
Renewable energy systems
Green building
Smart grid
a b s t r a c t
Renewable energy sources (RES) are an indigenous environmental option, economically competitive
with conventional power generation where good wind and solar resources are available. Hybrid systems
can help in improving the economic and environmental sustainability of renewable energy systems to
fulfill the energy demand. The aim of this paper is to present a dynamic model able to integrate different
RES and one storage device to feed a Green building for its thermal andelectrical energy needs in a sus-
tainable way. The system model is embedded in a dynamic decision model and is used to optimize a quite
complex hybrid system connected to the grid which can exploit different renewable energy sources. A
Model Predictive Control (MPC) is adopted to find the optimal solution. The optimization model has been
applied to a case study whereelectric energy is also used to pump waterfor domestic use. Optimal results
are reported for two main cases: the presence/absence of the energy storage system.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
The sustainable green building energy supply lead both devel-
oped and developing countries to make and implement new poli-
cies to improve efficiency in energy consumption, and to adopt
new alternatives like RES. Variable energy demands, intermittent
availability of renewable resources, different technological alterna-
tives to satisfy the different demands, and the possibility of inte-
grating storage and energy production systems give rise to the
exigency of defining criteria and strategies able to improve effi-
ciency and energy supply, and its environmental and economic
sustainability. Methods and models able to optimize such hybrid
systems for the specific case of a building are also welcomed be-
cause of the relevance that might have a Green building in terms
of environmental sustainability and energy efficiency.
In fact, in the EU and US, energy consumption in buildings has
even exceeded the energy consumption of the industrial and trans-
portation sectors [1]. Energy consumption of buildings accounts for
around 2040% of all energy consumed in advanced countries.
Over the last decade, more and more global organizations are
investing significant resources to create sustainable built environ-
ments, emphasizing sustainable building renovation processes to
reduce energy consumption and carbon dioxide emissions [1].
RES exploitation is one of the most important aspects of green
buildings. RES are defined as sources of energy that can be derived
from natural processes, and that can be replenished constantly, e.g.
energy generated from sun, wind, biomass, geothermal, hydro-
power [2]. The wind and solar energies are freely available and
environment friendly. The wind energy systems may not be tech-
nically viable at all sites because of low wind speeds and/or higher
unpredictability with respect to solar energy. Moreover, the avail-
ability of a specific resource depends on the specific season and
varies during the day. Integrating different RES in a hybrid system,
allowing their combined exploitation, is therefore becoming
increasingly attractive[3].
Hybrid renewable energy systems are becoming popular for re-
mote areapower generation applications dueto advances in renew-
able energy technologies and rise in prices of petroleum products.
Economic aspects of these technologies are sufficiently promising
to include them in developing power generation capacity [4]. Effec-
tive energy management of hybrid energy systems is necessary to
ensure optimal energy utilization and energy sustainability to the
maximum extent[5]. Hybrid systems can be considered as a rea-
sonable solution, capable to support systems that cover the energy
demands of both stand-alone and grid connected consumers that
can be integrated into residential, commercial, or institutional
buildings and/or industrial facilities. Furthermore, the hybrid
renewable energy systems are often the most cost-effective and
reliable way to produce power as well as to attenuate fluctuations
in power produced, thereby significantly reducing energy storage
0196-8904/$ - see front matter 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.enconman.2012.05.017
Corresponding author. Address: TEER, CNRST BP 8027 NU Rabat, Morocco.
Tel.: +212 666892377.
E-mail address: [email protected](A. Ouammi).
Energy Conversion and Management 64 (2012) 351363
Contents lists available atSciVerse ScienceDirect
Energy Conversion and Management
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e n c o n m a n
http://dx.doi.org/10.1016/j.enconman.2012.05.017mailto:[email protected]://dx.doi.org/10.1016/j.enconman.2012.05.017http://www.sciencedirect.com/science/journal/01968904http://www.elsevier.com/locate/enconmanhttp://www.elsevier.com/locate/enconmanhttp://www.sciencedirect.com/science/journal/01968904http://dx.doi.org/10.1016/j.enconman.2012.05.017mailto:[email protected]://dx.doi.org/10.1016/j.enconman.2012.05.0178/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
2/13
requirements [4]. In literature, the majority of papers have been fo-
cusedon the sizing of hybrid renewableenergy systems. Paskaet al.
[6]presented their experience of the design, build and exploitation
of many hybrid power systems. The results show that the proposed
hybrid plants are a good way to have available sources of electricity
which optimize utilization of primary energy sources. Ashok [7]
discussed different system components of hybrid energy system
and developed a general model to find an optimal combination of
energy components for a typical rural community minimizing the
life cycle cost while guaranteeing reliable system operation. Ekren
and Ekren [8] optimized an autonomous PV/wind integrated hybrid
energy system with battery storage. Bakos and Tsagas[9]reported
the technical feasibility and economic viability of a hybrid solar/
wind grid connected system for electrical and thermal energy pro-
duction, covering the energy demand of a typical residence in the
city of Xanthi (Greece). Reichling and Kulacki [10] modeled the per-
formance of a hybrid windsolarpower plant. They found that add-
ing solar thermal electric generating capacity to a wind farm rather
than expanding with additional wind capacity provides costbene-
fit. Dagdougui et al. [11] presented an overall optimization problem
in connection with a dynamic systemmodel, which is controlled in
real time to satisfy the hourly variable electric, hydrogen, and water
demands. Ouammi et al.[12]presented a practical methodology to
support decision makers in the evaluation of the wind exploitation
on a territory and the sustainability of the installation of a wind
power plant. Teleke et al. [13] developed a control strategy for opti-
mal useof thebatteryenergystorage. Papaefthymiou et al. [14] pre-
sented the wind-hydro-pumped storage hybrid power station of
Ikaria Island. Karki et al. [15]developed a methodology for an en-
ergy limitedhydro plant and wind farmcoordination using a Monte
Carlo simulation technique considering the chronological variation
in the wind, water and the energy demand. Zhang et al. [16]pre-
sented a multistage stochastic mixed integer programming model
for power generation in a day-ahead electricity market. Hyndman
and Fan[17]proposed a new methodology to forecast the density
of long-term peak electricity demand. Glanzmann and Andersson
[18] investigated a decentralized optimal power flow control foroverlapping areas in power systems. Abbey and Jos[19] studied
the problem of energy storage system sizing for isolated wind-die-
sel power systems. In[20], authors presented a building automa-
tion system, where the demand-side management is fully
integrated with the buildings energy production system, which
incorporates a complete set of renewable energy production and
storage systems. Ruther et al. [21] assessed the potential of grid
connected, building integrated photovoltaic generation in the state
capital, Florianopolis, in south Brazil. Al-Salaymeh et al. [22]stud-
ied the feasibility of utilizing photovoltaic systems in a standard
residential apartment in Amman city in Jordan. Hoes et al. [23] pro-
posed a concept that combines the benefits of buildings with low
and high thermal mass by applying hybrid adaptable thermal stor-
age systems and materials to a lightweight building. Braun andRther[24]showed the role of grid-connected building-integrated
photovoltaic in reducing the load demands of a large and urban
commercial building located in a warm climate in Brazil.
The aim of this paper is to define a dynamic optimization model
that is able to support decisions related to the operational manage-
ment of energy supply for a green building. Indeed, the model
could also be used for a more general grid-connected microgrid
characterized by different energy production plants and storage
systems, and, unlike the majority of works reported in the litera-
ture, is related to a system already sized and to the optimal control
of the overall system according to time-varying demands and re-
source availability.
Theinnovationcarried out by this study consistsin consideringa
mixed renewable energy sources and a storage system to feed aGreen building for its thermal and electrical energy needs in a
sustainable way. The systemmodel is embedded in a dynamic deci-
sion model and is used to optimize a quite complex hybrid system
connected to the grid. The optimization model has been solved for
twomain cases: thepresence/absence of theenergystoragesystem.
Finally, unlike most contributions available in the literature, in this
paper, an overall optimization problem is defined in connection
witha dynamic system model, whichis to be controlled in real time,
on the basisof the (discrete time) observation of the systemstate.In
particular, specific state and control variables are defined to formal-
ize the dynamic hybrid system model, the objective function, and
the constraints of the optimization problem. The problem consid-
ered in this paper is quadratic with linear constraints. A Model
Predictive Control (MPC) approach is used in the problem formal-
ization and solutionin order to take into accountupdates in the sys-
tem state, forecasts of the energy flows from renewable sources,
and demand variations, which can be forecasted with some difficul-
ties. The approach followed in this paper is based on the repeated
formalization and solution of an optimization problem over a finite
horizon, on the basis of the current observed state and on the avail-
able forecasts of the quantities of interest.
2. The control architecture
Intelligent buildings have attracted lots of attention in recent
years. Recent papers [25,26]have focused on the inclusion in the
Building Automation System (BAS) of algorithms based on control
and optimization, with specific reference to a Model Predictive
Control (MPC) scheme. Specifically, in [25], attention is focused
on the heating system and state equations are formalized to de-
scribe the temperature behavior in the different buildings rooms
as a function of the external temperature and of the heat sources.
Instead, in[26], authors developed an operational control platform
for an intelligent building using a SCADA (Supervisory Control and
Data Acquisition) system to control temperature and luminosity in
huge-area rooms. Energy management systems require the inte-
gration and communication of different kinds of information. Thus,a specific communication network should be developed [27].
In this paper, a dynamic decision model for a green building is
described in detail. The developed model is thought as a part of a
general BAS based on a MPC controller that integrates plants, stor-
age systems, energy demands (for electricity, heat, and water
pumping), resource availability forecasts, through an appropriate
communication system.
Fig. 1reports the information flows necessary for the dynamic
decision model application. A general architecture that can host
such a dynamic decision model is reported in [26]. The optimiza-
tion package used in this paper can be integrated with a supervi-
sory system. Otherwise, the dynamic decision model can also be
implemented in the Matlab Software (like in [26]). The results of
the optimization algorithm can be sent to an actuator system thatgives commands to the thermal, electrical, and water systems (i.e.,
inputs to production plants, inverters, storage system, air condi-
tioning, water pumping, etc.).
The architecture is based on a centralized intelligence that re-
ceives forecasts for renewable resources availability, demands,
prices, and system state (i.e., level of charge of the storage system),
and, on the basis of an optimization model, gives commands to a
sub-set of the production plants, to the storage system, to the
pump, and to the connection between the building and the exter-
nal net. The commands are given to typical hardware (i.e., invert-
ers, switch with the external net, production plants, etc.) that can
be present in the building. Moreover, the central controller is sup-
posed to communicate with the local controllers present in the
production plants and in the storage system. The primary controlvariables are those entities over which the central controller may
352 H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
3/13
physically decide. That is: the energy flows from/to the storage, the
production level of the plants that can be controlled, the energy
flows from/to the net, the status on/off of the pump. Other decision
variables are then used in this work for the electrical system (for
example, the splitting of the energy coming from the wind tur-
bine). These decision variables describe the theoretical behavior
of the energy flows, and, in fact, they are linked to the primary con-
trol variables through, for example, the energy balances present in
the overall system. For example, in the following, a variable is used
to define the energy from wind turbine/PV that is sent to the stor-
age system. In the case there is a very specific electrical network
(i.e., each plant is connected directly to the storage through specific
lines) this energy flow is the result of the energy balance among
production, loads, storage. Otherwise, this energy flow is sent to
the buildings grid and what the storage receives is undistinguish-
able, that is, one cannot know from which plant the flow comes.
However, also in the latter case, these variables can be defined as
secondary control variables that are linked through the primary
control variables through a set of constraints. Finally, it is impor-
tant to note that the aim of this paper is not the one of defining
a complete simulation model for the building subsystems (i.e.,
the electrical subsystem, the thermal subsystem, the water subsys-
tem). For this reason, simplified equations have been used to de-
scribe the electrical grid and the thermal subsystem, as well as
the plants and the storage.
3. The system model
3.1. Notations
The following definitions will be used throughout this paper to
describe the system model:
EtX The output energy produced from the wind turbine
(wt), photovoltaic (pv), biomass (b) and flat plate
collector (fpc) [kWh], in time interval (t, t+ 1),
t= 0, . . ., T 1; indexXdesignates the renewable
energy system and belongs to the set [wt,pv, b,fpc]
EtY;h The energy provided from renewable energy systems
and electrical network (net) [kWh] for heating (the
building) in time interval (t,t+ 1); index Y belongs to
the set [wt,pv,b,fpc, net]
EtZ;e The energy provided from the renewable energy
system and the network [kWh] for electric supply
[kWh] in time interval (t,t+ 1); index Z belongs to the
set [wt,pv, net]
EtZ;p The energy from the renewable energy system and the
electrical network [kWh] used to supply electricity for
pumping water in time interval (t,t+ 1);
t= 0, . . ., T 1
EtW;net The surplus produced fromwt/pv [kWh] sent to the
electrical network in time interval (t, t+ 1);
t= 0, . . ., T 1; index W belongs to the set [wt,pv]~Etnet The overall energy that is sent to the network fromwt
andPVmodules [kWh] in time interval (t, t+ 1);
t= 0,. . .
, T
1Etnet The overall energy [kWh] that is taken from the
network in time interval (t, t+ 1)
EtZ;st The energy from wind turbine/PV that is sent to the
storage system [kWh] in time interval (t,t+1);
t= 0, . . ., T 1
Qtw The amount of potable water provided to the
household [m3/h] in time interval (t,t+ 1),
t= 0, . . ., T 1
CHte The energy provided from the storage system to the
electricity [kWh] in time interval (t, t+ 1),
t= 0, . . ., T 1
CHth The energy provided from the storage system to the
heating system [kWh] in time interval (t, t+ 1);
t= 0, . . ., T 1
CHtp The energy provided from the storage system to the
pumping [kWh] in time interval (t,t+ 1);
t= 0, . . ., T 1
CBt The level of storage system [kWh] at time instant
t, t= 0, . . ., T 1
3.2. Consumer profile
The hybrid energy generation system for a Green building
connected to the electrical network (that is considered in this
work) is reported inFig. 2. The proposed system is developed for
a household that has electrical, thermal and water pumping needs.
However, the systemis enough general to be exported to industrial
Fig. 1. The building MPC-based architecture.
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 353
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
4/13
and/or residential areas and/or microgrids with similar character-
istics. This hybrid system consists of photovoltaic modules (PV),
a wind turbine, a small biomass plant, a solar flat plate collector
(FPC), and a storage system, as well a connection with the electrical
network.
The optimization of such system aims to generate energy satis-
fying the dynamic demands (i.e., heating demand, electrical de-
mand, and domestic water demand), which correspond to real
data of a household (of six persons) in the Liguria region in winter
season. The FPC and biomass plant are exclusively used to ensure
heating demand. Either energy produced by the wind turbine or
the energy produced by the PV modules can be directly used to sat-
isfy a part of the electrical demand as well as the water demand
through pumping, and/or can contribute to supply the heating de-
mand. The energy surplus from wind turbine and the PV modules
can be sent to the storage system or/and to the network. The stor-
age system can receive electrical energy from PV modules and
wind turbine and can provide free energy for heating, electricity,
and pumping needs in cases of deficit in electricity. Furthermore,
the network connection offers the possibility to purchase electric-
ity in case of failure of the storage system. The overall interactions
between the components of the system are reported in Fig. 3.
In order to ensure the sustainability of the hybrid system, and
addressing the mismatch between the intermittencies of wind
and solar irradiation, the system presented in this paper integrates
both an internal storage system and a connection with the electri-
cal network.
The necessity of properly integrating the access to the storage
and the electrical network gives rise to an optimal control problem.
3.3. Energy from the wind turbine
The following model is used to simulate the electrical power
output of the wind turbine[28]:
Ptwt
0 vt
v
c
prvt2v2cv2rv
2c
vc6 vt 6 vr
pr vr6 vt 6 vf
0 vt vf
8>>>>>>>:
t 0; . . . ; T 1 1
where vt is the forecasted wind speed in time interval. It is worth to
mention that the wind speed is predicted by some meteorological
model and hence these predictions are retained as reliable ones.
Then, pr is the rated electrical power, vcis the cut in wind speed,
vr is the rated wind speed, and vfis the cut off wind speed.
In general, the wind speed measurements are given at a height
different than the hub height of the wind turbine. The following
equation is used to evaluate the wind speed at the desired height
[11,29]:
vt
v
t
data
lnHhub=z
lnHdata=z t0;. . .
; T1
2
where Hdatais the height of the measurement, Hhub is the hub height,
z is the surface roughness length, and vtdata is the forecasted wind
speed at the height of the measurements.
Thus,
Etwt PtwtDt t 0; . . . ; T 1 3
3.4. Energy from the PV modules
The power generated from PV modules can be calculated using
the following formula[30]:
P
t
pv SpvgpvpfgpcG
t
t0;. . .
; T1
4
where Spv is the solar cell array area, gpv is the module referenceefficiency, pfis the packing factor, gpcis the power conditioning effi-ciency andGt is the forecasted hourly irradiance, that is predicted
by some meteorological model and hence these predictions are re-
tained as a reliable ones.
Thus,
Etpv Ptpv:Dt t 0; . . . ; T 1 5
3.5. Energy from the biomass heating plant
Energy provided by the biomass heating plant depends on the
used biomass quantity ut
[m
3
] in time interval (t, t+ 1), the bio-mass volumetric mass (VM) [kg m3] (i.e., the ratio between the
dry mass [kg] and the volume [m3]), the lower heating value
(LHV) [MJ kg1]. The LHV assumes that the latent heat of vaporiza-
tion of water in the fuel and the reaction products is not recovered.
Then, it can be calculated once higher heating value (HHV) and
moisture content (MC) are known. The HHV is the total energy re-
lease in the combustion with all of the products at 273 K in their
natural state when water has released its latent heat of condensa-
PV modules
Wind turbine
Biomass unit
Flat plate
collector (FPC)
Electrical Network
Heating/cooling
Electricity
Water
Energy storage
system
Hybrid system
Green building
Fig. 2. Energy flows between the hybrid system and the green household.
354 H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
5/13
tion. In the considered work, the HHV is evaluated from the
basic data analysis of biomass. The biomass MC represents the
water amount present in the biomass and it can be expressed as
a percentage of the dry weight. As regards production plant, the
plant is supposed to operate at the maximum productivity level.
The following equation provides the plant developed energy Etb
[kWh]:
Etb fgbhLHVutVM t 0; . . . ; T 1 6
wheregbh is the plant efficiency and fis a conversion factor [kWh/MJ].
3.6. Energy from the flat plate collector
The useful thermal energy extracted from the water collector
depends on the instantaneous incident solar irradiation, the plate
area, and its efficiency. That is[31],
Etfpc gfpcAfpcGtDt t 0; . . . ; T 1 7
where gfpcis the efficiency of the solar flat plate collector,Afpc[m2] is
the area andGt is the forecasted hourly irradiance, that is predictedby some meteorological model and hence these predictions are re-
tained as reliable ones.
3.7. Thermal, electrical and water needs
The hourly energy that can be used in time interval ( t,t+ 1) for
heating,Eth, is expressed by:
Eth Etwt;h E
tpv;h E
tfpc;h E
tb;h E
tnet;h CH
th t 0; . . . ; T 1
8
The hourly electrical energy, Ete, that can be used in timeinterval
is expressed by:
Ete Etwt;e E
tpv;e E
tnet;e CH
te t 0; . . . ; T 1 9
The hourly energy Etp that can be used in time interval for
pumping water is given by:
Etp Etwt;p E
tpv;p E
tnet;p CH
tp t 0; . . . ; T 1 10
The amount of pumped water is proportional to the energy used
for this purpose, that is:
Qtw
Etpqgh
gps t 0; . . . ; T 1 11
The energy that is sent to the electrical network is composed by
the surplus produced by the wind turbine and the surplus pro-
duced by the PV modules, i.e.,
~Etnet Etwt;net E
tpv;net t 0; . . . ; T 1 12
The energy that is taken from the network is given by:
Etnet Etnet;h E
tnet;e E
tnet;p t 0; . . . ; T 1 13
The energy productions by the wind turbine and the PV mod-ules in each time interval can be used for different purposes: to
supply electrical demand, heating demand, water demand, sent
to the storage system or sold to the network. Thus, the following
equations hold:
Etwt Etwt;h E
twt;e E
twt;p E
twt;net E
twt;st t 0; . . . ; T 1 14
Etpv Etpv;h E
tpv;e E
tpv;p E
tpv;net E
tpv;st t 0; . . . ; T 1 15
3.8. The storage system state equation
The storage system works as an inventory for the electrical en-
ergy that can, in this way, be stored. Specifically, a state equationfor the storage system can be written. That is,
Flat platecollector (FPC)
Wind turbine
Biomass unit
PV modules
Heating
Pumping water
Electricity
Electrical
Network
tewtE ,
thwtE ,
thpvE ,
thfpcE ,
thBE , t
hnetE ,
tepvE ,
tpwtE ,
tppvE ,
tpnetE ,
tenetE ,
tnetpvE ,
tnetpvE ,
Storage
systemt
stwtE ,
t stpvE ,
thCH
tpCH t
eCH
Fig. 3. Architectural system interactions.
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 355
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
6/13
CBt1 CBt Etwt;st Etpv;st CH
tp CH
th CH
te t
0; . . . ; T 1 16
4. The optimization problem
The decision variables of the optimization problem are Etwt;h,
Etpv;h, Etfpc;h, Etb;h, Etnet;h, Etwt;e, Etpv;e, Etnet;e, Etwt;st, Etpv;st, CHte, CH
th, CH
tp, while
CBt are state variables.
The objective function (to be minimized) is characterized by the
sum of different terms that are properly weighted: the deviation
from the various demands (electrical, thermal, domestic water),
the energy that is taken from the net (i.e., the system has the objec-
tive not to depend from the net but only from the produced energy,
when possible) and the energy in the storage system (that should
be maximized during the optimization horizon and at the end of
the optimization horizon). That is,
ZXT1t1
Etwt;h Etpv;h E
tfpc;h E
tB;h E
tnet;h CH
th E
tDh
2
aXT1t1
Et
wt;e Et
pv;e Et
net;e CHt
e Et
De2
bXT1t1
Etwt;p Etpv;p E
tnet;p CH
tp
qgh gps Q
tDw
!2
uXT1t0
Etnet cXT1t0
CBt qCBT t 0; . . . ; T 1 17
wherea, b,u,c andq are weighting factors.The constraints of the optimization problem are represented by
equations reported in Section 3. Moreover, there is a constraint re-
lated to the storage system. That is,
CBt 6 Cmax t 0; . . . ; T 1 18
whereCmax is the size of the storage system.However, obviously, the energy sent to the network must be
less than the sum of the energy produced by the wind turbine
and photovoltaic module.
~Etnet Etwt E
tpv t 0; . . . ; T 1 19
Instead, as regards biomass and flat collector plant, the pro-
duced energy can be less or equal to the available potential. In fact,
biomass may not be used (also because if it is summer, heating
does not work and heating cannot be sent to the network), and
water for heating passing through the plate collector may be
stopped. This implies the following relations
Etb;h Etb t 0; . . . ; T 1 20
Etfpc;h E
tfpc t 0; . . . ; T 1 21
The problem is here solved using mathematical programming
techniques through a commercial optimization package [32]. In
particular, the optimization problem is solved for two specific case
studies: presence/absence of the storage system. In the latter case,
thus, the variables related to the storage system are known and set
equal to zero.
5. Application to case study
The proposed dynamic decision model has been tested for the
Capo Vado site, which is the windiest one of Liguria region, in Italy
[33]. It is assumed that the forecasted data from which the optimi-
zation problem must be solved are exactly equal to the historicaldata recorded in the site, and which consist of the hourly wind
speed, recorded at the height of 10 m, and the hourly solar irradi-
ance (seeFig. 4).
The wind model described by Eqs. (1)(3)has been applied to
the specific case study, using the wind turbine G-3120 35 kW.
The parameters assume the following values: vc= 3.5 [m/s],
vr= 11 [m/s], vf= 25 [m/s], Pr= 35 [kW], Hhub= 30.5 [m],
Hdata= 10[m], z0= 0.03 [m]. For the PV modules, it is assumed
Spv
= 100 [m2],gpv
= 0.11,gpc
= 0.86, andPf
= 0.9.
Fig. 5 shows the hourly energy that is produced by the wind tur-
bine, and the PV modules. In fact, the figure demonstrates that the
wind energy is ranged between 0.35 kWh at 1:00 and 17 kWh at
19:00. However, for the solar modules, the energy output reaches
its maximum, 4.85 kWh, at 12:00. It is evident that the energy pro-
duction coming from the wind turbine is higher than the energy
produced from the PV modules. This fact is mainly due to the high
wind speed and the low solar radiation in the winter season.
5.1. Optimal results: absence of the storage system
In this case, additional constraints are introduced: all state and
control variables related to the storage system are set equal to
zero. Moreover, the related state equation has been eliminated.Figs. 6 and 7 show the obtained optimal results. Specifically,
Fig. 6shows how electrical energy is produced, whileFig. 7is re-
lated to the energy that is necessary for the pump (to satisfy the
domestic water demand).
According toFig. 6, the mainly part of the energy dedicated to
satisfy the electrical energy demand, approximately 66% comes
from the wind turbine, 18% comes from the PV modules, whereas
16% is taken from the electrical network (grid) between 1:00 and
7:00, this behaviour is mainly attributed to either the low wind
speed, absence of solar radiation and non-existence of storage
system.
Fig. 7 shows the hourly energy consumed by the pump, the
mainly part of this energy which is equal to 54% comes from the
wind turbine, and 30% of energy needs are supported by the elec-trical grid, while small part, less than 16% comes from the PV
modules.
The total energy surplus sold to the electrical network is re-
ported in Fig. 8. One can note that the contribution of the wind tur-
bine is dominated during all the day with a percentage of 94%,
while, a small part is sold from the PV modules between 9 h and
15 h which constitutes only 6% of the energy sold, this fact is due
to the low irradiance on winter season.
5.2. Optimal results: presence of the storage system
The optimal results to satisfy electrical energy needs of the
household are shown inFig. 9.
Approximately 55% results from wind turbine, 18% derives fromPV modules, 15% from the storage system and 12% from the grid.
The electrical energy demand reaches its higher hourly peak value
of 2.4 kWh at 18:00 and a minimum of 1.3 kWh at 04:00. The
hourly periods have been defined throughout the day: period A
(16 t< 7), period B (76 t< 15) and period C (156 t6 24). During
period A, the electrical energy needs were totally provided by the
wind turbine and the electrical grid, this combined contribution
is owing to both, low wind speed and absence of the solar radia-
tion, in addition, within this paper, it is assumed that initially the
storage system is empty. During the period B, and apart from time
7:00, the demand has been satisfied by various energy flows com-
ing from PV, wind turbine and storage system.
Finally, in the time period C, the electrical energy demand is en-
sured by both the wind turbine and storage system with dominantcontribution from wind energy.
356 H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
7/13
0
1
2
3
4
5
6
7
8
9
10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Windspeed(m/s)
0
100
200
300
400
500
600
Solarirradiance(W/m)
Wind-10 m
Wind-hub heightsolar
Fig. 4. Hourly wind speed and solar irradiance of the considered representative day at Capo Vado site.
0
2
4
6
8
10
12
14
16
18
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Energyproduced(kWh)
Wind
Solar
Fig. 5. Hourly energy produced.
No storage system
0
0,5
1
1,5
2
2,5
3
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Electricalenergyflow(kW
h)
Grid
Solar
Wind
Fig. 6. Electrical energy management (absence of storage system).
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 357
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
8/13
No storage system
0
2
4
6
8
10
12
14
16
18
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Energysenttothepump(kWh)
Grid
Solar
Wind
Fig. 7. Energy demand of pump (absence of storage system).
No storage system
0
2
4
6
8
10
12
14
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Energysoldtothenetwork(kWh)
Solar
Wind
Fig. 8. Energy sold to the electrical network (absence of storage system).
With storage system
0
0,5
1
1,5
2
2,5
3
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Electricalenergyflow
(kWh)
Storage system
Grid
Solar
Wind
Fig. 9. Optimal electrical energy control (presence of storage system).
358 H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
9/13
With storage sysytem
0
2
4
6
8
10
12
14
16
18
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Energysenttoth
epump(kWh)
Storage system
Grid
Solar
Wind
Fig. 10. Energy demand of pump (presence of storage system).
With storage system
0
2
4
6
8
10
12
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time (h)
Energysenttothestoragesystem(kWh) Solar
Wind
Fig. 11. Energy sent to the storage system (presence of storage system).
With storage system
0
5
10
15
20
25
30
35
40
45
50
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Storagesystemlevel(kWh)
Fig. 12. Storage system level.
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 359
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
10/13
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
11/13
5.3. Environmental impacts & benefits
One interesting issue is here related in understanding how
many greenhouse gas (GHG) emissions are avoided through the
use of a green building.
To this end, it is necessary to compare the emissions produced
by the green building and the emissions that would have been pro-
duced using conventional combustibles.
As regards CO2, the emissions produced by the green building
are zero, except when electrical energy from the main grid is used.
To calculate this quantity, the emission factor of the Italian electri-cal mix (0.465 kg CO2/kWhel) reported in[34]has been used.
Emissions generated by fossil fuel combustibles have been
calculated for two combustibles: coal and natural gas. For CO2,
the following equation is adopted:
COt2c fe;cfo;cQ
tD;c c 1;2 t 0; . . . ; T 1 22
where fe;c;fo;c are respectively the emission and oxidation factors,
that can be used when combustible c,c= 1,2, is used (and reportedin[34]), andQtD;cis the combustible flow of kind c, that needs to be
0
0,5
1
1,5
2
2,5
3
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Electricalenerg
y(kWh)
Provided
Demand
Fig. 15. Electrical energy satisfaction.
0
2
4
6
8
10
12
14
16
18
20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
CO
2reduction(kg)
Coal (best practice)
Natural gas (CCGT)
Fig. 16. Hourly CO2 gas reduction.
Table 1
Daily GHG reductions.
Energy sources CO2(kg) SO2(kg) NOx (kg)
Coal (BP) 180 2.2 0.8
Natural gas (CCGT) 81 0.09
BP: Best Practice.CCGT: Combined Cycle Gas Turbine.
Table 2
GHG emissions (absence/presence of storage system).
Energy sources Green-house gases emission (no storage system)
CO2 (kg) SO2 (kg) NOx (kg)
Panel A
Coal (BP) 39 0.48 0.18Natural gas (CCGT) 17 0.02
Green-house gases emission (with storage system)
Panel B
Coal (BP) 20 0.25 0.09
Natural gas (CCGT) 9.3 0.01
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 361
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
12/13
used to satisfy the household energy demand in time interval
(t, t+ 1). This flow has been calculated using average data relevant
to the average characteristics of plants present in Italy [34].
The avoided emissions are finally calculated subtracting this
quantity to the emissions that could be generated by fossil fuels.
Fig. 16shows the CO2 reduction by using the proposed hybrid
energy system for a household instead of producing the house-
holds needs by the conventional fossil fuel (coal and natural gas).
Similar considerations can be derived for other GHG gas emis-
sions (i.e., SO2 and NOx, in this work). The only difference is that
there is not an emission factor of the Italian electrical mix. In this
case, data related to the typical plants present in Italy have been
used.
In Table 1, the GHG reductions are shown during the considered
day. As notated in the previous sections, the hybrid system is con-
nected to the electrical network, so part of energy demand is satis-
fied by the grid which is supposed to be a non-renewable energy
system. Thus, an assessment of the GHG emissions should be done.
Table 2reports the evaluation of the GHG emitted for both system
configurations (presence/absence of the storage system). It can be
summarized that large amount of GHG can be avoided by coupling
the storage system to the hybrid energy system. Specifically,
approximately 50% of the GHG emissions are reduced. The environ-
mental impacts have been reduced significantly by implementing
the storage system, which, in fact, enhance the reliability of the
proposed green hybrid energy system.
Other impacts related to the different management options and
in particular related to the costs of the purchased electrical energy
can be calculated. It is important to note that, in this paper, a spe-
cific objective related in the problem formalization to costs has not
been included. This means that the optimal solution does not find
the minimum costs related to policies of purchasing/sales of the
electrical energy. However, an interesting analysis is related to
the daily cost of the electrical energy taken from the net in case
where the storage is present or not. Specifically, considering the
optimal results reported inFigs. 5, 6, 8, and 9, and taking the elec-
trical energy prices from data available during the day for the Ital-ian electrical market, the following conclusions can be drawn for
one day:
If the storage is not present, the overall daily cost is equal to
6.5.
If the storage is present, the overall daily cost is equal to 3 .
6. Conclusion
A dynamic decision model for the energy management of a
household has been proposed. The model is able to define the opti-
mal energy flows management in a building characterized by a mix
of renewable energy resources (solar, biomass, wind) to satisfy dif-
ferent demands (electric, heating and water). A dynamic decisionmodel has been developed and applied to a household located at
Capo Vado (Liguria Region). Optimal results to satisfy all the energy
demands are found for a testing day with presence/absence of the
storage system. Moreover, an analysis about the avoided emissions
is reported and the importance of having a storage system is dem-
onstrated also under this point of view. The developed decision
model can also be used for real time management, through a
MPC approach that is adopted for the dynamic decision model
solution.
It should be noted that, in the case of absence of the energystor-
age system, since there are no state equations and available energy
varies in each time interval, the optimization problem can be run
at each time interval in a separated way, and is in fact static.
Future developments of the present work regard the possibilityof introducing stochastic issues both in demand and resource pre-
dictions. In this case, in order to reduce the overall complexity of
the problem, dynamic programming could be used. Then, the mod-
el could be detailed from the electrical, thermal and technological
point of view. Moreover, an objective function based on energy
prices could also be adopted in order to consider the green-build-
ing as an energy producer.
Moreover, a multi-objective approach could be formalized to
take into account economic and environmental issues, as well as
the objectives weighted in the current version of the objective
function.
Finally, the inclusion of the developed dynamic decision model
in a decision support system that can allow the user to interact
with the mathematical models could be helpful for the energy
management of the household.
Acknowledgments
Authors wish to thank ARPAL (Agenzia Regionale per la Protezi-
one dellAmbiente Ligure), that has kindly collected and made
available the important historical data set at the basis of this work.
References
[1] Juan YK, Gao P, Wang J. A hybrid decision support system forsustainableoffice
building renovation and energy performance improvement. Energy Build
2010;42:2907.
[2] Xia C, Zhu Y, Lin B. Renewable energy utilization evaluation method in green
buildings. Renew Energy 2008;33:8836.
[3] Nema P, Nema RK, Rangnekar S. A current and future state of art development
of hybrid energy system using wind and PV-solar: a review. Renew Sustain
Energy Rev 2009;13:2096103.
[4] Deshmukh MK, Deshmukh SS. Modeling of hybrid renewable energy systems.
Renew Sustain Energy Rev 2008;12:23549.
[5] Pourmousavi SA, Nehrir MH, Colson CM, Wang C. Real-time energy
management of a stand-alone hybrid wind-microturbine energy system
using particle swarm optimization. IEEE Trans Sustain Energy 2010;1(3).
[6] Paska J, Biczel P, Klos M. Hybrid power systems an effective way of utilising
primary energy sources. Renew Energy 2009;34:241421.
[7] Ashok S. Optimised model for community-based hybrid energy system. Renew
Energy 2007;32:115564.
[8] Ekren O, Ekren BY. Size optimization of a PV/wind hybrid energy conversion
system with battery storage using response surface methodology. Appl Energy
2008;85:1086101.
[9] Bakos GC, Tsagas NF. Technoeconomic assessment of a hybrid solar/wind
installation for electrical energy saving. Energy Build 2003;35:13945.
[10] Reichling JP, Kulacki FA. Utility scale hybrid windsolar thermal electrical
generation: a case study for Minnesota. Energy 2008;33:62638.
[11] Dagdougui H, Miniciardi R, Ouammi A, Robba M, Sacile R. A dynamic decision
model for the real time control of hybrid renewable energy production
systems. IEEE Syst J 2010;4(3):32332.
[12] Ouammi A, Sacile R, Zejli D, Mimet A, Benchrifa R. Sustainability of a wind
power plant: application to different Moroccan sites. Energy
2010;35:422636.
[13] Teleke S, Baran ME, Bhattacharya S, Huang AQ. Rule-based control of battery
energy storage for dispatching intermittent renewable sources. IEEE Trans
Sustain Energy 2010;1(3).
[14] Papaefthymiou SV, Karamanou EG, Papathanassiou SA, Papadopoulos MP. A
wind-hydro-pumped storage station leading to high res penetration in theautonomous island system of Ikaria. IEEE Trans Sustain Energy 2010;1(3).
[15] Karki R, Hu P, Billinton R. Reliability evaluation considering wind and hydro
power coordination. IEEE Trans Power Syst 2010;25(2):68593.
[16] Zhang J, Fuller JD, Elhedhli S. A stochastic programming model for a day-ahead
electricity market with real-time reserve shortage pricing. IEEE Trans Power
Syst 2010;25(2):70313.
[17] Hyndman RJ, Fan S. Density forecasting for long-term peak electricity demand.
IEEE Trans Power Syst 2010;25(2):114252.
[18] Glanzmann GH, Andersson G. Decentralized optimal power flow control for
overlapping areas in power systems. IEEE Trans Power Syst
2009;24(1):32736.
[19] AbbeyC, Jos G. A stochastic optimization approach to ratingof energystorage
systems in wind-diesel isolated grids. IEEE Trans Power Syst
2009;24(1):41826.
[20] Figueiredo J, Martins J. Energy production system management renewable
energy power supply integration with building automation system. Energy
Convers Manage 2010;51:11206.
[21] Ruther R, Knob PJ, Jardim CS, Rebechi SH. Potential of building integrated
photovoltaic solar energy generators in assisting daytime peaking feeders inurban areas in Brazil. Energy Convers Manage 2008;49:10749.
362 H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363
8/11/2019 Modeling and Optimization of a Hybrid System for the Energy Supply
13/13
[22] Al-Salaymeh A, Al-Hamamre Z, Sharaf F, Abdelkader MR. Technical and
economical assessment of the utilization of photovoltaic systems in residential
buildings: the case of Jordan. Energy Convers Manage 2010;51:171926.
[23] Hoes P, Trcka M, Hensen JLM, Bonnema BH. Investigating the potential of a
novel low-energy house concept with hybrid adaptable thermal storage.
Energy Convers Manage 2011;52:24427.
[24] Braun P, Rther R. The role of grid-connected, building-integrated photovoltaic
generation in commercial building energy and power loads in a warm and
sunny climate. Energy Convers Manage 2010;51:245766.
[25] Siroky J, Oldewurtel F, Cigler J, Privara S. Experimental analysis of model
predictive control for an energy efficient building heating system. Appl Energy2011;88:307987. http://dx.doi.org/10.10116/j.apenergy.2011.03.00.
[26] Figueiredo J, Sa da Costa J. A SCADA system for energy management in
intelligent buildings. Energy Build 2012. http://dx.doi.org/10.10116/
j.enbuild.2012.01.041.
[27] Neumann P. Communication on industrial automation what is going on?
Control Engineering Practice 2007;15(11):133247. http://dx.doi.org/
10.10116/j.conengprac.2006.10.00.
[28] Notton G, Muselli M, Poggi P, Louche A. Decentralized wind energy system
providing small electrical loads in remote areas. Int J Energy Res
2001;25:14164.
[29] Ouammi A, Dagdougui H, Sacile R, Mimet A. Monthly and seasonal assessment
of wind energy characteristics at four monitored locations in Liguria region
(Italy). Renew Sustain Energy Rev 2010;14:195968.
[30] HocaogluFO, GerekON, Kurban M. A novel hybrid (windphotovoltaic) system
sizing procedure. Solar Energy 2009;83:201928.
[31] El Fadar A, Mimet A, Perez-Garca M. Modelling and performance study of a
continuous adsorption refrigeration system driven by parabolic trough solar
collector. Solar Energy 2009;83:50861.[32] Lingo 10, .
[33] Ouammi A, Sacile R, Mimet A. Wind energy potential in Liguria region. Renew
Sustain Energy Rev 2010;14:289300.
[34] Caputo A. Thermoelectric energy production and CO2 emission. ISPRA; 2011.
H. Dagdougui et al. / Energy Conversion and Management 64 (2012) 351363 363
http://www.lindosystems.org/http://www.lindosystems.org/