ASME-ATI-UIT 2015 Conference on Thermal Energy Systems: Production, Storage, Utilization and the Environment

17 – 20 May, 2015, Napoli, Italy

MULTI-OBJECTIVE OPERATIONAL OPTIMIZATION OF A DISTRIBUTED

ENERGY SYSTEM FOR A LARGE-SCALE UTILITY CUSTOMER

Marialaura Di Somma1,2, Bing Yan3, Nicola Bianco1, Peter B. Luh3,

Giorgio Graditi2, Luigi Mongibello2, Vincenzo Naso1
1
Dipartimento di Ingegneria Industriale (DII) – Università degli Studi Federico II – Napoli – 80125 – Italy
2
ENEA - Italian National Agency for New Technologies, Energy and Sustainable Economic Development - CR
Portici – Portici – 80055 – Italy
3
Department of Electrical and Computer Engineering – University of Connecticut – Storrs – CT 06269 – USA.

ABSTRACT
With energy saving issues and growing environment protection awareness, interest in distributed generation has been
intensifying. Distributed Energy Systems (DESs) are being widely investigated, since they are expected to be largely used to
increase the efficiency of energy supply and to address environmental problems. In this paper, a multi-objective optimization
problem is formulated to obtain the optimal operational strategies of a DES, to reduce both the energy cost and the environmental
impact. The DES includes different energy conversion devices and thermal energy storages to satisfy time-varying user demands.
The Pareto front, including the best possible trade-offs between the economic and the environmental objectives, is obtained by
minimizing a weighted sum of the total energy costs and CO2 emissions. The operators of DESs can choose the operational
strategy from the Pareto front based on the economic and environmental priorities. The model is implemented for a DES for a
large-scale utility customer. Results show that the optimized operation of the DES reduces energy costs and CO2 emissions, as
compared with conventional energy supply systems. In addition, a sensitivity analysis is carried out to analyze the effects on
energy costs and environmental impact of variations in the configuration of the DES supply side.

Keywords: Distributed Energy System, Total energy cost and Cooling Heat and Power (CCHP) system, solar thermal plant,
CO2 emissions, Multi-objective operational optimization. reversible heat pump and thermal energy storages, which
provide electricity, heat and cooling to end-users. The
1. INTRODUCTION economic objective is formulated as the total energy cost to be
minimized, and the environmental objective is formulated as
A Distributed Energy System (DES) may consist of small- the total CO2 emissions to be minimized. The Pareto front
scale heat and power generation technologies including also involving the best possible trade-offs between the economic
renewable ones, and storage units, providing electric and and environmental objectives is attained by minimizing a
thermal energy to end-users. In recent years, DESs have been weighted sum of the total energy cost and CO2 emissions, by
recognized as an effective alternative to conventional energy using branch-and-cut. The operators of DESs can choose the
supply systems [1, 2], since they allow to increase the operational strategy from the Pareto front based on the
efficiency of energy supply as well as to address economic and environmental priorities.
environmental issues. However, most of the studies in the As an illustrative example, a large-scale utility customer (a
literature have been focused on the optimized operation of large hotel located in Italy) is considered as the end-user.
specific energy systems, such as Combined Heat and Power Results show that the optimized operation of the DES allows
(CHP) systems, from the economic point of view [3-5]. to reduce the energy costs and the environmental impact, as
The economic analysis alone is not sufficient due to compared with conventional energy supply systems. In
growing environmental concerns, like the global warming and addition, a sensitivity analysis is carried out, to analyze the
the depletion of fossil fuels. Operation problems of DESs, effect on energy costs and environmental impact of variations
including different energy systems, become more challenging in the configuration of the DES supply side.
when the environmental aspects are also taken into account,
since the economic and the environmental objectives may be 2. PROBLEM FORMULATION
contradictory [6]. In addition, the energy devices involved
convert and store different energy carriers (electricity, natural The DES under consideration consists of energy
gas, solar energy, hot and/or cold fluids) with different energy conversion devices and thermal energy storages, providing
efficiencies and environmental impacts. electricity, heat and cooling to end-users. Figure 1 shows the
In this paper, a multi-objective linear programming scheme of the DES with the possible routes of energy carriers
(MOLP) problem is formulated to obtain the optimal from various energy resources via primary and secondary
operational strategies of a DES to reduce the energy costs and energy devices, and thermal storages to meet given time-
the environmental impact, while satisfying time-varying user varying user demands. Modeling of energy devices and
demands, with given prices of energy sources. The DES thermal storages is presented in Subsection 2.1, energy
involves different energy conversion devices: Combined balances are described in Subsection 2.2.
 ASME-ATI-UIT 2015 Conference on Thermal Energy Systems: Production, Storage, Utilization and the Environment
17 – 20 May, 2015, Napoli, Italy

Figure 1. Scheme of the DES for the optimization problem

Q ICE,ex (t )  E CCHP (t )1  e   ICE  / e ,

2.1. Modeling of Energy Devices and Thermal Storages
(3)
The common constraint for most of the energy devices is
the capacity constraint, formulated as follows: where µ ICE is the percent heat loss of the engine.

 kED (t )  xED (t )kED

Engine exhaust gas can be subdivided among the heat
min max
xED (t )kED , (1) recovery boilers and the absorption chiller to supply heating
for the domestic hot water and space heating demands as well
that means that if the energy device is in use (i.e., the on/off as cooling for the space cooling demand.
binary decision variable, xED(t), is 1), its generation level (the The heat rate supplied by the exhaust gas to the heat
decision variable), kED(t), has to be within the minimum recovery boiler for the domestic hot water demand,
min
value, k ED max
, and its capacity k ED , and 0 when the energy H ex
DHW
(t ) , is:

(t )  Q ICE,ex (t ) DHW (t ) HR,boil ,

device is off.
Additional constraints for CCHP, solar thermal plant, H ex
DHW
(4)
reversible heat pump and thermal storage are presented in the
following. where ηHR,boil is the efficiency of the heat recovery boiler and
2.1.1. Modeling of the CCHP system. The CCHP system the continuous decision variable, ξDHW (t), is the fraction of
consists of an internal combustion engine (ICE); two heat exhaust gas supplied to the heat recovery boiler for the
recovery boilers, for Domestic Hot Water (DHW) and Space domestic hot water demand.
Heating (SH) demands, respectively; an absorption chiller, for Heating can be also directly provided by supplementary
Space Cooling (SC) demand, as sketched inside the bold lines burning of natural gas in the heat recovery boiler. The
in Figure 1. The internal combustion engine provides volumetric flow rate of natural gas, G boil
DHW
(t ) , required by the

 
electricity fuelled by natural gas. Thermal energy is recovered 
boiler to directly provide the heat rate, H DHW
(t ) , is:
di
from exhaust gas and used to provide heating by the heat
recovery boilers and cooling by the absorption chiller. G boil
DHW
(t )  H di
DHW
(t ) / boilLHVgas , (5)
Furthermore, heating and cooling can be also directly
generated by a supplementary burning of natural gas in the where ηboil is the combustion efficiency of the boiler.
boilers and absorption chiller, respectively [7]. Decision Therefore, the total generation of the heat recovery boiler
variables for the CCHP system are the electricity generation
for the domestic hot water demand, H CCHP
DHW
(t ) , is the sum of
level in the internal combustion engine, the fraction of
exhaust gas for domestic hot water, space heating and cooling the heat rate obtained by exhaust gas, H  DHW
(t ) , and the heat
ex
demands, and heating and cooling directly provided by the rate directly provided by supplementary burning of natural
supplementary burning of natural gas in the boilers and gas, H di
DHW
(t ) :
absorption chiller, respectively.
Constraints considered for the CCHP system are presented H CCHP
DHW
(t )  H ex
DHW
(t )  H di
DHW
(t ) . (6)
below. The volumetric flow rate of natural gas, G ICE (t ) ,
Modeling of heating for the space heating demand and of
required by the engine to provide electricity, E (t ) , is
CCHP cooling for the space cooling demand by the CCHP system is

 
given by: similar to that described above.
G ICE (t )  ECCHP (t ) / e LHVgas ,
The sum of the engine exhaust gas fractions used for
(2) domestic hot water, ξDHW(t), space heating, ξSH(t), in the heat
recovery boilers, and space cooling, ξSC(t), in the absorption
where ηe is the engine gas-to-electric efficiency and LHVgas is chiller, has to be one:
 DHW (t )   SH (t )   SC (t )  1 .
the lower heat value of natural gas.
The heat rate available from the exhaust gas recovered (7)
from the engine, Q ICE ,ex (t ) , is:
 The overall volumetric flow rate of natural gas consumed 2.2.2. Domestic hot water energy balance. The heat rate
by the CCHP system, G CCHP (t ) , is: demanded for domestic hot water, H dem DHW
(t ) , must be

GCCHP (t )  GGT (t )  Gboil (t )  Gboil (t )  G abs(t ) ,

DHW SH
satisfied by the total heat rate provided by the CCHP system,
(8)
H CCHP
DHW
(t ) , by the solar thermal plant, H solar (t ) , and by the
where GboilSH
(t ) is the volumetric flow rate of natural gas thermal storage, H out (t )  H in (t ) , that is:
sto sto

(t )  H CCHP (t )  H solar(t )  H sto (t )  H sto

required by the boiler to directly provide heating for the space
H dem
DHW DHW out in
(t ) . (13)
heating demand, and Gabs (t ) is the volumetric flow rate of
natural gas required by the absorption chiller to directly The space heating and cooling balances can be expressed
provide cooling for the space cooling demand. in a similar way.
2.1.2. Modeling of the solar thermal plant. A solar thermal
3. MULTI-OBJECTIVE OPTIMIZATION
plant is used to meet the domestic hot water demand. The heat
rate provided by the solar plant, H solar (t ) , depends on the The objective is to minimize the total energy costs and the
collector area, Acoll, its efficiency, ηcoll, and the total solar CO2 emissions. The economic and environmental objective
irradiance, G T (t ) , and is expressed by: functions are discussed in Subsections 3.1. To solve the

H solar(t )  coll AcollGT (t ) ,

problem, the multi-objective optimization method is discussed
(9) in Subsection 3.2.

where the collector area is assumed to be known, since the 3.1. Economic and Environmental Objectives
optimal design of the energy devices is not the aim of this
work. The economic objective is to minimize the total energy
2.1.3. Modeling of the reversible heat pump. A reversible cost, Cost, that is the cost of the gas consumed by the CCHP
heat pump is used to meet space heating and cooling demands system, G CCHP (t ) , and the cost of buying electricity from the
in the heating and cooling modes, respectively. In the heating
mode, the electricity consumption of the heat pump, E HP (t ) , power grid, E buy (t ) :

 P t ,
to supply the heat rate, H (t ) , is given by:
Cost  grid (t ) Ebuy (t )  PgasGCCHP (t )
HP
 
E HP (t )  H HP (t ) / COPHP ,
(14)
(10) t

where COPHP is the coefficient of performance of the heat where Pgrid(t) is the time-of-day unit price of electricity from
pump in the heating mode. Modeling of cooling mode is the grid and Pgas is the constant unit price of natural gas.
similar to that described above. The environmental objective is to minimize the
2.1.4. Modeling of the thermal energy storages. The environmental impact, Env, in terms of CO2 emissions from
energy stored in the domestic hot water tank at time t, Hsto(t), the power grid and the consumed fuels. The CO2 emissions

 
can be expressed as: due to the use of electricity from the power grid are evaluated

H sto (t )  H sto (t  t )sto  H sto (t )  H sto (t ) t ,

in out
by multiplying the carbon intensity of the power grid, Ecin,
(11)
and the total amount of electricity from the grid, E (t ) . The buy
where ηsto is the efficiency of the thermal storage and Δt is the carbon intensity of the power grid is the amount of CO2
length of the time interval. The decision variables are H sto
in
(t ) emissions per unit of electricity generated by the specific
and H (t ) , which are the heat rates brought in and taken out
out
sto
power grid. It depends on the fuel mix used to generate
electricity for the power grid at which the DES is connected.
by the flow-in and flow-out water, respectively.
The CO2 emission due to the natural gas consumption is
It is assumed that there are three different thermal energy
evaluated by multiplying the carbon intensity of the fuel, Gcin,
storage systems, each of them for the corresponding thermal
and the total amount of fuel consumption of the CCHP
energy demand. Modeling of thermal storage systems for

E t .
space heating and cooling is similar to the above. system, G CCHP (t ) , [8]:

2.2. Modeling of Energy Balances Env  cin Ebuy(t )  GcinGCCHP (t )

  (15)
t

In order to satisfy the given time-varying user demands, 3.2. Multi-Objective Optimization Method
electricity and thermal energy balances are formulated by
matching supply and demand. With the economic objective function (Eq.14) and the
2.2.1. Electricity balance. The electricity rate demand, environmental one (Eq.15), the problem has two objective
E dem (t ) , and the electricity rate required by the heat pump, functions to be minimized. To solve this multi-objective
E (t ) , must be covered by the sum of the electricity rate problem, a single objective function is formulated as a
HP
weighted sum of the total energy cost, Cost, and the
delivered by the CCHP system, E CCHP (t ) , and the electricity environmental impact, Env, to be minimized:

Fobj  cCost  1   Env ,

rate bought from the grid (a continuous decision variable),
(16)
E (t ) :
buy

E dem (t )  E HP (t )  ECCHP (t )  Ebuy(t ) . (12)

where the constant c is chosen such as the order of magnitude
of the terms c Cost and Env is the same. For ω = 1, the
economic optimization is carried out and the solution that
minimizes the total energy cost can be found. For ω = 0, the
environmental impact optimization is carried out and the
solution that minimizes the total CO2 emissions can be found.
Then, the constant c is calculated as the ratio of the maximum
total CO2 emissions obtained by the economic optimization to
the maximum total energy cost obtained by the environmental
impact optimization. With the constant c, the Pareto front
involving the best possible trade-offs between the two
objectives can be found by varying the weight ω in the range
0 – 1. The above formulated problem is linear and involves Figure 3. DES configuration.
both discrete and continues variables. This Mixed Integer
Linear Programming (MILP) problem is solved by branch- 4.1.2. Energy prices. The time-of-day unit price of
and-cut. Figure 2 shows the flow-chart to find the optimal electricity from the power grid and the unit price of natural
operational strategies of the DES, accounting for both gas are chosen according to the current Italian market
economic and environmental objectives. Given the input data, scenario. Reference is made to the Italian BTA6 tariff [12] for
by solving the above MOLP problem, the Pareto front, the electricity from the power grid. The tariff structure for
consisting of the best possible trade-offs between the two industrial use considered in this study is shown in Figure 5.
objectives, can be obtained. Each point of the Pareto front Reference is made to a unit price (€/kWh) of the electricity
corresponds to a different operational strategy of the DES. from the power grid that is the sum of the energy and
The operators of DESs can choose the operational strategy dispatching prices, the power distribution and transmission
from the Pareto front based on the economic and quotas, the equalization component and the excise fee. For the
environmental priorities. natural gas, the tariff for industrial use is adopted [13].
Reference is made to a unit price (€/Nm3) consisting of the
4. NUMERICAL TESTING energy quotas (energy unit price and additional charges), the
other variable quotas as distribution and transport sale quotas,
The above formulated problem has been implemented by and the excise fee.
using IBM ILOG CPLEX Optimization Studio Version 12.5. 4.1.3. Primary energy carriers. The energy carriers input to
As an illustrative example, a hypothetical large hotel of the DES are electricity from the power grid, natural gas and
16,000 m2 located in Italy (D climatic zone) is considered as solar energy. The first two are assumed unlimited, whereas
the targeted end-user. A typical winter day is chosen, with one the heat rate provided by the solar thermal plant is derived by
hour as time-step. The configuration of the DES, including the solar energy input taken from meteorological data for the
the sizes of energy devices, is sketched in Figure 3. considered location [14]. The hourly solar irradiance of a
representative winter day is evaluated as the average of the
4.1. Model inputs solar irradiance in the corresponding hours of all winter days.
4.1.4. Carbon intensity. Information about the carbon
The required inputs for the optimization model are demand intensity of electricity from the power grid and natural gas are
information, energy prices information, primary energy needed to evaluate the total amount of CO2 emissions related
carriers availability, carbon intensity information, and to the operation of the DES. The carbon intensity of the power
efficiencies of energy devices and storages, as discussed in grid is assumed equal to 0.354 kg/kWh, as the averaged value
the following. in the years 2009-2011 for Europe [15]. The carbon intensity
4.1.1. Energy demand. The hourly electricity, domestic of natural gas is assumed equal to 0.202 kg/kWh [8].
hot water and space heating demands are taken from the 4.1.5. Efficiency of energy devices and thermal storages.
literature [9-11]. The energy rate demand profiles for a typical Typical values of the efficiency assumed for the energy
winter day are reported in Figure 4. devices and thermal storages are reported in Table 1. The
temperature of exhaust gas from the internal combustion
engine is assumed equal to 623.15 K and the temperature of
exhaust gas at the exit of the heat recovery boiler is assumed
equal to 363.15 K [8]. The efficiency of the heat recovery

Figure 2. Flow chart of the multi-objective optimization Figure 4. Energy rate demands of a hypothetical hotel in Italy.
model.
Figure 5. Time-of-day electricity unit price for industrial use Figure 6. Pareto fronts with and without discount on the
according to the Italian BTA6 tariff. excise fee of natural gas.

boiler is evaluated as the ratio of difference between the inlet In the second case (with discount on the excise fee of
and the outlet temperature of the engine exhaust gas in the natural gas), the Pareto front is obtained in the same way,
heat recovery boiler to the difference between the inlet where the points marked with a' and b' are obtained by
exhaust gas temperature and the ambient temperature. minimizing the daily energy cost and the daily CO2 emissions,
respectively.
4.2. Results In both cases, a significant reduction, 8%, in the CO2
emissions is gained from solution points a and a' (ω = 1) to
The results are presented in the following. the solution points c and c' (ω = 0.9) with a negligible 0.25%
4.2.1. Pareto front. Figure 6 shows the Pareto fronts obtained increase in the energy cost. The differences between the two
without and with the discount on the excise fee of natural gas Pareto fronts become more significant when the weight for
applicable to high-efficiency cogeneration systems (Primary the economic objective increases (left side). In the energy cost
Energy Saving (PES) > 0) [16]. In the first case, the natural minimization, at the point a', the daily energy cost is 1,223
gas tariff for industrial use described in Subsection 4.1.2 is €/d and it is reduced by about 3% as compared with the
adopted, considering the excise fee for industrial use for all energy cost at the point a. The total CO2 emissions are the
the natural gas consumed by the CCHP system [17]. In the same as those at the point a. When the weight for the
second case, the discount on the excise fee for natural gas is environmental objective increases, the sensitivity of the DES
involved, since the CCHP system has a PES > 0. According operation to the energy prices reduces, therefore the
with [17], this discount is applied to a 0.25 Nm3 volumetric difference between the Pareto fronts reduces (right side). At
flow rate of natural gas consumed by the CCHP system for the point b', the daily energy cost is 1,589 €/d and the daily
each kWh of electricity provided. The additional consumption CO2 emissions are the same as those at the point b, since,
of natural gas, which occurs when the CCHP system has an when the environmental objective is minimized, the operation
electrical efficiency less than 42%, is subjected to the of the DES is not sensitive to the energy prices. In the
industrial excise fee. Also the natural gas consumed by the environmental impact minimization, the daily energy cost at
boilers to directly provide heating for the domestic hot water the point b' is almost equal to that at the point b, because of
and space heating demands is subjected to the industrial the very small difference between the discounted excise and
excise fee [17]. the full excise prices for industrial use.
In the first case (without discount on the excise fee of 4.2.2. Optimal operational strategies at various trade-off
natural gas), the point marked with a is obtained by points. Each point on the Pareto front corresponds to a
minimizing the daily energy cost, and the daily energy cost is different operational strategy of the DES. The operators of the
1,260 €/d, whereas the daily CO2 emissions are 4,160 kg/d. DES can choose a compromise between the two objectives
The point marked with b is obtained by minimizing the from the Pareto front based on their cost and environmental
environmental impact (the daily CO2 emissions), and the daily priorities. In order to understand how the operational
energy cost is 1,594 €/d, whereas the daily CO2 emissions are strategies of the DES affect the energy cost and the CO 2
3,448 kg/d. The points between the extreme points are found emissions under different weight values, the results at various
by equally subdividing the weight interval into 100 spaces. trade-off points are shown in Figure 7. These trade-off points
belong to the Pareto front obtained when the discount on the
Table 1. Efficiency of energy devices and thermal storages. excise fee is involved (red Pareto front in Figure 6). Figure
7A points out that, as ω increases from 0 to 1, the share of the
Efficiency electricity load (the sum of electricity demand and electricity
Primary energy devices
Electrical Thermal rate required by the heat pump) satisfied by the CCHP
Internal combustion engine 0.35 0.50 significantly increases (5% to 59%), highlighting that the
Solar thermal plant 0.40 CCHP system allows to reduce the total energy cost. The
Secondary energy devices Efficiency opposite occurs to the share of electricity load covered by the
Heat pump COPHP = 3.0 grid power. The maximum value is obtained when the
Heat recovery boiler ηHRboil = 0.75 ηboil = 0.85 environmental impact is minimized, since the space heating
Thermal energy storage Efficiency demand is fully satisfied by the heat pump, as shown in Figure
DHW and SH storages 0.90 7c. As ω increases from 0 to 1, the share of the space heating
 increases, and the integration with the boiler driven by natural
gas is not needed. When the weight of the economic objective
is close to 1 (ω = 0.8 and ω = 0.9), the natural gas boiler is
used to satisfy a small share (3 - 5%) of the domestic hot
water demand. Although in the economic optimization the use
of the CCHP system attains the maximum value, exhaust gas
are not enough to satisfy the domestic hot water and space
heating demands. Therefore, the use of natural gas boilers
significantly increases consistently with the reduction in the
use of the heat pump. The remarkable difference in the
operation of the DES from ω = 1 to ω = 0.9 corresponds to
the big jump from a' to c' shown in Figure 6.

4.3. Sensitivity Analysis

A sensitivity analysis is carried out to analyze the effects

on energy costs and environmental impact of variations in the
configuration of the DES supply side.
4.3.1. Single-objective optimization for different
configurations of the DES supply side. In order to show the
contribution of each energy device in reducing energy costs
and CO2 emissions separately, the economic and
environmental impact optimizations are carried out for
different configurations of the DES supply side. In addition,
the daily energy cost and CO2 emissions are also evaluated for
one of the most common conventional energy supply systems,
consisting of the power grid to meet the electricity demand,
and natural gas boilers to meet the domestic hot water and
space heating demands. The configurations are listed in Table
2.
Figures 8A and B show the daily energy cost obtained by
the economic optimization and the daily CO 2 emissions
obtained by the environmental impact optimization,
respectively, for the above listed configurations.
Configuration 1 is the reference case, consisting of all energy
devices shown in Figure 3. In the reference case (discount on
the excise fee of natural gas involved), the minimum energy
cost and CO2 emissions are obtained by the economic and
environmental impact optimization, respectively. For
configuration 2, there is a negligible increase in the energy
cost and no change in CO2 emissions, compared with the
minimum energy cost and the minimum CO2 emissions of the
reference case. This highlights the small impact of the space
heating storage on both the objectives.
Configuration 3 excludes the solar thermal plant. A 3%
Figure 7. Optimal operational strategies of the DES at various increase in the energy cost, and a 5% increase in the CO2
trade-off points for A) Electricity, B) Domestic Hot Water, C) emissions compared with the minimum energy cost and the
Space Heating. minimum CO2 emissions of the reference case, respectively,
confirm the importance of this energy device for both the
demand satisfied by the heat pump decreases, whereas the
share covered by the heat recovery boiler driven by exhaust Table 2. Investigated configurations.
gas increases because of the increase in the use of the CCHP
system. In the energy cost minimization, the operation of the Energy devices excluded from the
Configuration
DES is only sensitive to energy prices, and the CCHP system reference case (Configuration 1)
instead of the power grid is mostly used to provide electricity. 2 without SH storage
3 without solar thermal plant
It is also worth noting that Figures 7A, 7B and 7C are 4 without solar thermal plant and DHW storage
strongly related. For instance, the CCHP system is rarely used without solar thermal plant,
in the environmental impact minimization, since the space 5
DHW/SH storages
heating demand is fully satisfied by the heat pump. 6 without heat pump
Correspondingly, the heat rates from exhaust gas and solar 7 without internal combustion engine
thermal plant do not satisfy the domestic hot water demand, Configuration Other cases
and the integration with the boiler driven by natural gas is Gas turbine instead of internal combustion
8
needed. As the use of the CCHP system increases and the use engine
of the heat pump reduces, the amount of exhaust gas 9 Conventional energy supply system
 This confirms that the internal combustion engine is a better
solution in terms of costs and environmental impact than the
gas turbine, due to the higher total energy conversion
efficiency of the CCHP system with the internal combustion
engine than that of the CCHP system with the gas turbine. It
can be noticed that the effect of changing the prime mover is
larger on energy costs than on the environmental impact.
Finally, for the conventional energy supply system, the
daily energy cost and the daily CO2 emissions are 27% and
26% higher than the minimum energy cost and the minimum
CO2 emissions, respectively, in the reference case. Results
show that the energy cost and the environmental impact are
strongly reduced by the optimized operation of the DES.
4.3.2. Multi-objective optimization for different
configurations of the DES supply side. The multi-objective
optimization is carried out for some configurations among
those listed in Table 2, in order to compare the Pareto fronts
with that obtained in the reference case (red Pareto front in
Figure 6). The Pareto fronts for Configurations 3, 6 ,8, as well
as for the reference case are presented in Figure 9.
For Configuration 3 (without solar thermal plant), the
Pareto front is similar to that in the reference case, especially
in the left side (ω is close to 1). The CO2 emissions are highly
reduced (10%) from ω = 1 to 0.9, with a negligible increase in
the energy cost (0.31%). When the weight of the
environmental objective increases over that of the economic
one (right side), the slope of the Pareto front is lower than that
of the reference case. This means that when more attention is
paid to the environmental performance, the rate of the energy
Figure 8. A) Daily energy cost for Configurations 1-9 in the costs increase is larger than that of CO2 emission reduction, as
economic optimization; B). Daily CO2 emissions for compared with the reference case. The daily energy cost
Configurations 1-9 in the environmental impact optimization. obtained by the economic and environmental impact
optimization are 24% and 22% lower than those obtained with
objectives. Besides the solar thermal plant, Configurations 4 the conventional energy supply system (Configuration 9).
and 5 exclude the domestic hot water storage and both the For Configuration 6 (without heat pump), very few trade-
thermal storages respectively, and results are similar to those offs points are obtained. The difference between the
obtained for configuration 3. maximum and minimum daily energy cost obtained by the
Configuration 6 excludes the heat pump. The daily energy environmental impact and economic optimization,
cost is 11% higher than the minimum energy cost in the respectively, is only 1.5%. The difference between the
reference case. The daily CO2 emissions are 21% higher than maximum and minimum daily CO2 emissions obtained by the
the minimum CO2 emissions in the reference case. Therefore, economic and environmental impact optimization,
the heat pump affects the environmental impact more than the respectively, is only 0.34%, because of the flat shape of the
energy costs, as was also shown in Figure 7C. In the Pareto front, which implies a negligible reduction in CO 2
environmental impact optimization, the space heating demand emissions with a significant increase in the energy cost. Daily
is fully satisfied by the heat pump, whereas in the economic energy costs obtained by the economic and environmental
optimization the share of space heating demand satisfied by impact optimization are 18% and 7% lower than those
the heat pump attains the minimum value. obtained with the conventional energy supply system
Configuration 7 excludes the internal combustion engine.
The electricity load is fully satisfied by the power grid and,
without exhaust gas, the heat recovery boilers are fuelled by
natural gas. The opposite trends of energy costs and CO2
emissions as compared with those of Configuration 6 are
exhibited. The daily energy cost is 25% higher than the
minimum energy cost in the reference case. However, the
daily CO2 emissions are only 0.5% higher than the minimum
CO2 emissions in the reference case. This is also remarkable
in Figure 7A, since in the environmental impact optimization
only 5% of the electricity load is satisfied by the CCHP
system.
In Configuration 8, the internal combustion engine is
substituted by a gas turbine of the same size. The energy cost
is increased by 5% and the CO2 emissions are increased by
0.3%, as compared with the minimum energy cost and the Figure 9. Pareto fronts for Configurations 3, 6, 8 and
minimum CO2 emissions in the reference case, respectively. Reference case.
(configuration 9). [6] H. Ren, W. Zhou, K. Nakagami, W. Gao, Q. Wu. Multi-
For configuration 8 (gas turbine generator instead of objective optimization for the operation of distributed energy
internal combustion engine), the shape of the Pareto front is systems considering economic and environmental aspects,
different from that in the reference case. The slope of the Applied Energy 87 (12) (2010) 3642-3451.
Pareto front changes more quickly than that of the Pareto [7] X.Q. Kong, R.Z. Wang, X.H. Huang. Energy optimization
front in the reference case. When ω changes from 1 to 0.9, for a CCHP system with available gas turbines, Applied
there is a negligible increase in the energy cost (0.12%), and Thermal Engineering 25 (2005) 377-391.
the reduction in the CO2 emissions is about 3.4%. The daily [8] Educogen, The European Educational Tool on
energy costs obtained by the economic and environmental Cogeneration, Second Edition, December 2001.
impact optimization are 23% and 26% lower than those [9] M. Aprile. Caratterizzazione del settore alberghiero in
obtained with the conventional energy supply system Italia, Report ENEA RSE/2009/162 (2009).
(configuration 9). [10] M. Beccali, M. La Gennusa, L. Lo Coco, G. Rizzo. An
empirical approach for ranking environmental and energy
5. CONCLUSIONS savings measures in the hotel sector, Renewable Energy 34
(2009) 82-90.
In this paper, a MOLP problem is formulated to obtain the [11] L. Mongibello, N. Bianco, M. Caliano, G. Graditi, M.
optimal operational strategies of a DES to reduce both the Musto. Optimal operation of micro-CHP systems for a single-
energy cost and the environmental impact, while satisfying family house in Italy, Applied Mechanics and Materials, 492,
given time-varying user demands. The Pareto front, consisting (2014) 467-472.
of the best possible trade-offs between the daily energy cost [12] Delibera Autorità per l’Energia Elettrica e per il Gas
and CO2 emissions, is obtained by minimizing a weighted 119/11.
sum of the economic and environmental objectives by using [13] Autorità per l’Energia Elettrica e per il Gas
branch-and-cut. Results show that the operation of the DES is http://www.autorita.energia.it/it/prezzi.htm.
sensitive to the weights assigned to the two objectives. The [14] ASHRAE International Weather files for Energy
operators of DESs can choose the operational strategy from Calculations (IWEC weather files). Users manual and CD-
the Pareto front based on their cost and environmental ROM, American Society of Heating, Refrigerating and Air-
priorities. The contributions of each energy device in reducing Conditioning Engineers, Atlanta, GA, USA, 2001.
energy costs and CO2 emissions are evaluated by a sensitivity [15] IPCC. Guidelines for national greenhouse gas
analysis. In addition, the Pareto fronts for different inventories - CO2 emissions from fuel combustion, highlights,
configurations of the DES supply side are also discussed. The 2013 Edition, IEA (2013).
optimized operation of the DES in the reference case allows [16] Guida alla Cogenerazione ad Alto Rendimento CAR, 1st
to maximize the reduction in terms of costs (27%), and CO2 Edition, GSE (2012).
emissions (26%) as compared with conventional energy [17] Delibera Autorità per l’Energia Elettrica e per il Gas
supply systems. 16/9.

ACKNOWLEDGMENTS

Authors thank the Università di Napoli Federico II for

funding this study within the agreement with the University of
Connecticut and the Smart grid con sistemi di poligenerazione
distribuita (Poligrid).

REFERENCES

[1] K. Alanne, A. Saari. Distributed energy generation a

sustainable development, Renewable Sustainable Energy
Reviews 10 (6) (2006) 539-558.
[2] S.R. Allen, G.P. Hammond, M.C. McManus. Prospects for
and barriers to domestic micro-generation: a United Kingdom
perspective, Applied Energy 85 (6) (2008) 528-544.
[3] A.D. Hawkes, M.A. Leach. Cost-effective operating
strategy for residential micro-combined heat and power,
Energy 32 (5) (2007) 711-723
[4] M. Houwing, A.N. Ajah, P.W. Heijnen, I. Bouwmans,
P.M. Herder. Uncertainties in the design and operation of
distributed energy resources: the case of micro-CHP systems,
Energy 33 (10) (2008) 1518-1536.
[5] T. Wakui, R. Yokoyama, K. Shimizu. Suitable operational
strategy for power interchange operation using multiple
residential SOFC (solid oxide fuel cells) cogeneration
systems, Energy 35 (2) (2010) 740-750.