This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Energy Conversion and Management 49 (2008) 3006–3016

Contents lists available at ScienceDirect

Energy Conversion and Management

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

EABOT – Energetic analysis as a basis for robust optimization

of trigeneration systems by linear programming
A. Piacentino a,*, F. Cardona b
a
DREAM – Department of Energetic and Environmental Researches, Università di Palermo, Viale delle Scienze, 90128 Palermo, Italy
b
DINCE – Department of Nuclear Engineering and Energy Conversions, Univ. ‘‘La Sapienza’’, Cso. Vitt. Emanuele II, 244 Rome, Italy

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

Article history: The optimization of synthesis, design and operation in trigeneration systems for building applications is a
Received 12 November 2007 quite complex task, due to the high number of decision variables, the presence of irregular heat, cooling
Accepted 18 June 2008 and electric load profiles and the variable electricity price. Consequently, computer-aided techniques are
Available online 6 August 2008
usually adopted to achieve the optimal solution, based either on iterative techniques, linear or non-linear
programming or evolutionary search. Large efforts have been made in improving algorithm efficiency,
Keywords: which have resulted in an increasingly rapid convergence to the optimal solution and in reduced calcu-
Trigeneration
lation time; robust algorithm have also been formulated, assuming stochastic behaviour for energy loads
Optimization
Linear programming
and prices. This paper is based on the assumption that margins for improvements in the optimization of
Thermoeconomics trigeneration systems still exist, which require an in-depth understanding of plant’s energetic behaviour.
Multi-objective Robustness in the optimization of trigeneration systems has more to do with a ‘‘correct and comprehen-
sive” than with an ‘‘efficient” modelling, being larger efforts required to energy specialists rather than to
experts in efficient algorithms. With reference to a mixed integer linear programming model imple-
mented in MatLab for a trigeneration system including a pressurized (medium temperature) heat storage,
the relevant contribute of thermoeconomics and energo-environmental analysis in the phase of mathe-
matical modelling and code testing are shown.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction to understand the complexity of the optimization problem, the fol-

lowing aspects can be remarked:
In the last decade, a growing interest has been observed for
combined heat and power (CHP) or combined heat, cooling and – Safety of supply and flexibility are usually ensured by redun-
power (CHCP) applications in buildings. This is obviously due to dancy, i.e. the system is designed as a ‘‘facility of systems of
the high conversion efficiency of polygeneration systems and the a same product”, where different components may alterna-
consequent energy and pollutant-emissions savings, but also due tively contribute to cover the demand of a specific energy
to favourable external conditions, like the new opportunities exist- vector.
ing in the liberalised energy market and the growth of a ‘‘small – The problem is time-dependent: the variability in energy
scale CHP” market, which has gradually reduced the purchase loads and prices requires the adoption of flexible plant oper-
and installation cost of CHP units (typically, in the order of 600– ation strategies; in the civil sector, discretization on hourly
800 €/kWe). basis is usually pursued, resulting in a high number of deci-
These promising perspectives have stimulated the efforts of sci- sion variables as concerns plant operation. Discussions have
entists towards the definition of criteria for the optimization of arisen on the possibility to adopt a reduced set of ‘‘standard
CHCP design and operation for applications in the civil sector. Sev- days” (typically defined on seasonal and ‘‘working–non-
eral analyses have been oriented to assess the potential benefits in working” bases) without loss of reliability, but this is a con-
terms of energy and pollutant-emissions savings [1,2] and, in some troversial argument which needs ad hoc considerations for
cases, some peculiar aspects were examined adopting thermoeco- each case.
nomic cost-accounting methods [3] or pinch analysis [4]. In order – The decision variables have a non-homogeneous nature,
both as concerns the way they affect the objective function
and the values they can assume. Either in case of profit,
* Corresponding author. Tel.: +39 091 236 302; fax: +39 091 484 425.
energy or pollutant-emission saving-oriented optimizations,
E-mail address: [email protected] (A. Piacentino). the objective function depends on annual results, calculated

0196-8904/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2008.06.015
 Author's personal copy

A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016 3007

Nomenclature

a, b constants in linearized equations TES thermal energy storage

B generic flux (energy or exergy terms) Z, z total and specific capital cost
C cooling demand or production
CHCP combined heat, cooling and power Superscripts
CHP combined heat and power abs absolute
CONS energy consumption conv conventional
COP coefficient of performance dir for direct uses
D energy load, on hourly basis rel relative
E electricity demand or production
FOR feasible operation range Subscripts
GA genetic algorithms Abs absorption chiller
H heat demand or production boilmax corresponding to boiler capacity (heat peak load)
i* interest rate del delivering
LL load level el.ch. electric chiller
LP, NLP linear and non-linear programming Exch exchanged with the grid
LR Lagrangian relaxation inv investment
m mass of hot water stored pow.plant power plant
MP market price ret return
NPV net present value
PES primary energy saving Greek symbols
PHR power to heat ratio d 0–1 binary variable
STOR stored energy (kWh) g efficiency
T temperature

as sum of single values obtained for each time-step. The level have been assumed. This aspect heavily influences the choice
optimization problem can be divided into three different of the most appropriate resolution technique. Evolutionary search,
sub-problems: for instance, which has been extensively used in the optimization of
(a) Synthesis: in order to optimize the plant configuration, energy systems by genetic algorithms (GA), is not suitable for our
i.e. to select what components should be installed, a problem because of the deficiency in handling highly constrained
starting redundant ‘‘superconfiguration” is usually problems; also, GA could be preferably adopted to optimize plant
adopted, that is a general CHCP scheme where several operation as internal routine of an iterative synthesis-design opti-
components are included and a high level of intercon- mization [5] and this approach is not suitable for the examined
nections among them is assumed. The decision vari- problem due to the huge number of different operating conditions.
ables at synthesis level are 0–1 variables, each Several heuristic approaches have been proposed, oriented to
indicating the decision to include/not include a cer- determine near-optimal solutions basing on ‘‘aggregate thermal
tain component. load” duration curves [6] or thermoeconomics [7]; most of them,
(b) Design: CHCP systems for buildings applications are however, do not include an ‘‘integrated” optimization process, but
usually made up by highly standardised components assume a priori a sub-optimal management strategy (either ‘‘heat
(gas turbines, reciprocate engines, water–lithium bro- tracking” or ‘‘electricity tracking” operation modes). More recently,
mide absorption chillers, etc.), which can be regarded linear programming (LP) techniques have been extensively used
as black boxes and modelled by defining their part [8,9], due to the possibility to solve large scale problems with thou-
load behaviour. The absence of variables involving sands variables approaching the ‘‘multi-level” optimization prob-
the thermodynamic state of working fluids (the opti- lem by an ‘‘horizontal algorithm”, where synthesis, design and
mization of heat exchangers represents a 2nd refine- operational variables are threaten similarly. More refined ap-
ment level, not considered in this paper) makes the proaches have been proposed in [10], where a robust optimization
design optimization easier than usual. Only a rela- included a sensitivity analysis in LP to consider stochastic energy
tively small number of design variables representing loads, and in [11], where an efficient algorithm was proposed,
the size of plant components is included. which resulted much faster than an efficient sparse simplex code.
(c) Operation: the optimization of plant operation is more The fact that the production of the three energy vectors follows
complex than usual; in fact, CHCP systems offer sev- a joints characteristic makes often convenient to include thermal
eral possibilities of loading the different components energy storages (TESs, i.e. hot water and/or chilled water tanks);
to cover energy requests, the optimal solution usually, the TES is used to maximise power production during peak
depending on efficiency figures, energy loads and hours (where high value electricity is produced), storing eventual
prices. This optimization level involves both 0–1 vari- surplus heat/cooling energy to reuse it during off-peak hours.
ables (the on/off state for each component) and con- The inclusion of a TES significantly varies the structure of the opti-
tinue variables (the load level of each component). mization problem, introducing dynamic constraints; a clear over-
Also, the optimization routine must be applied on view of the techniques proposed in the literature to deal with
hourly basis, because both energy load and prices storage constraints was provided in [12]. Let us here briefly resume
are time-dependent. the two main currencies:

The variables of the different sub-problems are not of a ‘‘same – Decoupling the time-dependent storage constraint, a set of
rank”; for instance, operational variables could be optimized only small-size single-period sub-problems may be solved. In
once fixed values for the decision variables at synthesis and design [12,13] Lagrangian relaxation (LR) methods were used,
 Author's personal copy

3008 A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016

which decompose the original problem into multiple under- considering a generic building, its requests for heat, cooling and
lying Lagrangian sub-problems. In [13], the unit commit- electricity (respectively, DH, DC and DE) are assumed to be known
ment (where the on/off state of each component is on hourly basis. From a qualitative point of view, typical 6 °C/
determined) and the economic dispatch (the optimal pro- 13 °C delivering/return temperature are supposed for the cold
duction rates) are solved by isolating the economic dispatch water to/from the air conditioning units; only low temperature
problem (a unit commitment is assumed known, which is in (up to 90 °C) heat requests for space heating and sanitary hot water
its turn optimized by LR) and relaxing the time-coupling are considered, which could be covered by both low and high tem-
constraint obtaining a dual problem which does not require perature heat recovery in reciprocate engines. The same determin-
dynamic programming. istic approach is assumed for energy prices; accordingly to the
– Treating the multi-period large-size problem as unique and current trend in the European Union, where net-metering mecha-
solving it by Simplex or interior point methods require lar- nisms are gradually extended to new power producers, a same MPE
ger computational efforts, but a less complicate problem price for electricity purchase/sell is assumed, which is obviously
formulation. time-dependent. In the reference scheme assumed as CHCP super-
configuration, no reversible heat pump will be included, but a low
The first of the two above currencies has been attracting more temperature thermal energy storage (TES).
interest among researchers and the optimization of trigeneration
systems has gradually become a research field for experts in oper- 2.1. Modelling plant components
ational research. Innovative methods frequently converge to opti-
mal solutions whose energetic consistence might be said The optimization model proposed in this paper includes both
controversial, as evident when considering that a same lay-out, opti- economical and energo-environmental analyses; hence, appropri-
mized by different algorithms and adopting a same set ‘‘objective ate cost and performance figures for the plant components must
function + constraints + boundary conditions”, frequently con- be introduced. Most CHCP components have significant scale-fac-
verges to quite different solutions [14]. In particular, the solutions tor in the ‘‘purchase and installation” cost; however, the cost of
achieved are often too sensitive to slight variations in the boundary the CHP unit can be approximated by a linear expression:
conditions, which reflects a non-comprehensive representation of
energetic conversion processes. Analytical CHCP modelling should Z CHP;inv ¼ z0;CHP  EaCHP;nom ffi aCHP;inv ECHP;nom þ bCHP;inv dCHP ð1Þ
be more closely interrelated to energetics, as may be observed in
[15] where reliability of results is ensured by an energo-environ- where ECHP,nom and dCHP, respectively, represent the electric capac-
mental analysis of plant performance. In this paper we put into evi- ity of the CHP unit and the 0–1 binary variable expressing the deci-
dence that only an in-depth energetic analysis may properly reflect sion to include/not include the unit. On more than 200 CHP units
the ‘‘formation” of the values of the objective function, which ulti- from different producers [17], this approximation introduced
mately drive any optimization algorithm towards a final solution. ±0.01 variations in the correlation coefficient R2. As concerns the
Hence, no enhanced efficiency algorithms are proposed; on the con- part load performance, a crucial aspect is now discussed, that is
trary, a common Linear-Programming Interior Point method (a var- the use of 0–1 binary variables for the on/off state of components.
iant of Mehrotra’s predictor–corrector method) was used, which is Part load efficiency penalty is usually kept into account by imposing
the default option for large scale optimization in MatLab. All the a ‘‘feasible operation range” for the jth component in the ith hour:
analyses that will be presented have been used to code a MatLab FORj;i ¼ ½dopj;i ¼ 0 [ ½ðdopj;i ¼ 1Þ \ ðLLj;i 2 ½LLmin ; LLmax Þ ð2Þ
tool, EABOT (Energy Analysis-Based Optimization of Trigeneration
plants), which represent a module of a larger-scale software for The limit expressed by Eq. (2) with a logic operator is usually
the optimization of CHCP-based l-grids [16] not completed yet. integrated in LP models as follows:
Ahead in the paper four main concepts are critically examined:
LLmin;j  dopj;i 6 LLj;i 6 LLmax;j  dopj;i ð3Þ
1. Is the definition of a few ‘‘reference days” (as concerns load pro- This apparently non-linear constraint does not influence the
files) necessary? At what extent selecting a small number of ref- solution search method, being dop-j,i a 0–1 binary variable. Qua-
erence days affects the reliability of results and reduces the dratic or linear ge(LL) equations are usually assumed as concerns
consumption of computational resources? part load efficiency; alternatively, linear expressions for the char-
2. Once all operational constraints have been properly defined, is acteristic equations g [18] can be assumed:
it possible to determine additional conditions based on
Binput ¼ g i ðx; Boutput Þ ¼ ai  LLi  Boutput;nom þ bi di ð4Þ
advanced energetic analyses in order to make the optimization
algorithm converge faster? which are more suitable for LP models. Below in the paper, varia-
3. Can multi-objective optimization offer a more relevant contrib- tions in the power to heat ratio (PHRCHP) are neglected and, as con-
ute in CHCP decision making? A pure profit-oriented optimiza- cerns the absorption chiller, a fixed 0.70 COPabs is assumed. Also, as
tion would not properly reflect all the implications related to concerns the auxiliary boiler and the vapour compression chiller,
polygeneration: the support mechanisms in force, like tax linear cost figures and constant gboil (0.90) and COPel.chill. (2.5) are
exemption or dispatching priority, and the recent legislation used; both new and existing customers can be considered by
for CHP eligibility give a particular focus on the energo-environ- including or not the capital cost of these two components in the
mental effects, which will become more and more profit- objective function expression. The consumption of resources in
conditioning. mixed integer linear programming (MILP) problems strongly de-
4. Does the testing phase of CHCP optimization tools require ener- pends upon the number of integer variables dop-i; optimizing plant
go-environmental analysis? operation on hourly basis, i.e. 8760 values per year, could become
very time-consuming. A simplified method is here adopted. Let us
2. Problem formulation assume LLCHP and LLabs variable in the whole range [0, 1] and neglect
efficiency drops at part load operation (i.e. constant ge); let us as-
CHCP applications in the civil sector can be considered high ris- sume not to introduce binary variables for the on/off state, a signif-
ky/high profit-potential investments, which require accurate via- icant reduction in the calculation time being expected. Is such a
bility analysis and energy audits to have been performed. Then, roughly approximate approach reliable? In order to answer this
 Author's personal copy

A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016 3009

question, we may observe that CHP heat recovery (cold production Synthesis and design
in the absorption chiller) may be considered as ‘‘alternative” to pro-
dCHP ; dAbs ; dTES ; ECHP;nom ; C Abs;nom ; V TES
duction of heat (cold) in a constant efficiency boiler (vapour com-
pression chiller). Consequently, in a certain hour, the optimal unit Operation
commitment and economic dispatch problem can be generalized
HCHP;i ; C Abs;i ; LLboil;i ; LLel:ch:;i ; STORi ; Q TES;i ; for i ¼ 1 to Nhours
as follows: a combined production unit 1 produces two outputs a
and b, which could alternatively be produced by components 2
and 3, respectively; let us assume a fixed a/b production ratio for 2.2.1. The objective function
the unit 1 and constant efficiencies g2 and g3. Minimizing an objec- Let us start with a brief analysis of the objective function. If a
tive function (OF, linear equation in Binput,1, Binput,2 and Binput,3): profit-oriented approach is adopted, appropriate economic indica-
tors will be needed. In [8] the payback period was assumed as deci-
OF ¼ const1 Binput;1 þ const2 Binput;2 þ const3 Binput;3 ð5Þ
sion function, resulting in a small-size CHP unit to cover the base
subject to load; in [15] a more reliable approach is used, keeping simulta-
neously into account different economic indicators capable to ex-
Da ¼ Baoutput1 þ Baoutput2 ¼ LL1 Baoutput1;nom þ LL2 Baoutput2;nom ;
ð6:a; bÞ ploit the full profit-potential through the plant life cycle. Among
Db ¼ Bboutput1 þ Bboutput2 ¼ LL1 Bboutput1;nom þ LL3 Bboutput3;nom them, the net present value (NPV) has been proven the most useful
for CHP decision making [20] and is here assumed as objective
is equivalent to impose null:
function in the profit-oriented optimization (actually, its opposite
a a
o Boutput1;nom Boutput2;nom ‘‘-NPV” is minimized):
OF ¼ const1   const2  (
oLL1 ga1 ga2 NX
hours
HCHP;i PHRCHP LLboil;i Eboil;nom

Bboutput3;nom min P=Aði ; nÞ  b  cCHP
gas þ cboil
gas
 const3  ð7Þ i¼1
gE;CHP gboil
gb3   
LLel:ch:
 HCHP;i  PHRCHP  Eel:ch:;nom  DE;i  cE
Unless the marginal cost (equivalent to the average cost, in the COPel:ch:
hypothesis of constant efficiencies) of the two alternative produc- þ aCHP;inv  ECHP;nom þ bCHP;inv  dsinth;CHP þ aabs;inv  C Abs;nom
tion methods are equal, Eq. (7) indicates OF to represent a mono- þ babs;inv  dsinth;Abs þ aTES;inv  V TES þ bTES;inv  dsinth;TES ð10Þ
tone function. Minimizing OF would consequently lead to assume
LL1 alternatively equal to 0 or Da =Baoutput1;nom ; if the combined pro- In Eq. (10), the possibility to assume different gas prices for feeding
duction unit 1 is no oversized, for most of the year the above de- the CHP unit and the boiler is kept into account; ‘‘P/A(i*,n)” repre-
scribed simplified approach will lead to a technically-feasible sents the actualization factor. The b factor, which is related to the
operation. This is of course a quite ‘‘heavy” simplification, whose possibility to fix an arbitrary number of time periods Nhours to be
reliability will require an ex post analysis. Finally, as concerns the used as a basis for the optimization, will be discussed in the next
short-term heat storage, for pressurized steel tanks in [19] the subsection.
prevalence of heat losses through the tank walls and due to the For an energy/pollutant-emissions saving-oriented optimiza-
contact between the hot and the cold water was pointed out. A tion, maximization of annual savings can be only achieved with
few energetic models are available in literature, following two reference to ‘‘absolute indicators”. In the conventional definition
main currencies: of the primary energy saving index PES%:
conv CHP
– Assumption of a fixed (usually, 8–10%) energy loss between Fuel  Fuel
PESrel
% ¼ conv ð11Þ
charging and discharging; Fuel
– Assumption of a fixed DHloss loss per hour, in percentage Fuelconv is defined with reference to the actual CHP production
(usually, 2–3%) of the currently stored energy. rate and significantly depends on whether the plant is a net power
seller/buyer along the year. Only when the annual energy saving is
The latter approach is here adopted, expressing the TES balance compared to a fixed energy consumption [15], maximising PES% is
by two variables, STOR (energy stored at each time-step) and HTES,i equivalent to maximise the energy saving. With reference to Fig. 1,
(charge/discharge rate at each time-step i): where the CHCP superconfiguration and its energy flows are com-
  pared to the ‘‘separate production” case, let us indicate with CON-
DHloss
STORTES;iþ1 ¼ 1  STORTES;i  HTES;i ð8Þ Sconv the energy consumption to cover the current loads by a
100
conventional system:
where  
DH 1 DC
CONSconv ¼ þ  DE þ ð12Þ
STORi ¼ mi  cp  ðT del  T ret Þ ð9Þ gboil gpow:plant COPel:ch:
Evidently, these energetical variables are related to the volume
The primary energy saving, expressed in kW (average value in a
VTES, that is the parameter most affecting the purchase and instal-
certain hour), is
lation cost. Again, Eq. (1) can be formulated assuming VTES as
!
capacity variable; a brief cost analysis suggested to assume aTES
conv 1 1
and bTES equal to 155 €/m3 and 2050 €, respectively. PES ¼ CONS  PHRCHP   HCHP
gE;CHP gpow:plant
2.2. MILP model: objective function and constraints Hboilmax DE DC;max
 LLboil   LLel:ch: ð13Þ
gboil gpow:plant COPel:ch: gpow:plant
In this section, the optimization problem is formulated accord-
ingly to the assumptions proposed in Section 2.1; a multi-objective The need for ‘‘absolute” energy saving indicators can be put into
approach is adopted to keep into account all economical and ener- evidence with reference to an example. Let us consider a hour
go-environmental implications. Let us assume the following set of where DE, DH and DC are, respectively, 70 kWe, 80 kWh and
decision variables: 310 kWc; let us also assume gpow.plant equal to 0.44. If we assume
 Author's personal copy

3010 A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016

DC,max LL el.ch.,i
Combined Heat, Cooling and Power COPel.ch. Auxiliary DC,max LL el.ch.,i
el. chiller

grid Cabs,i
DC Separate Production
D COPabs
PHR H CHP,i - D E - C,max LL el.ch.,i
COPel.ch. Absorption
chiller El. chiller
Cabs,i
PHR H CHP,i
ηe,CHP 1 DC
η pow.plant (D E +
COPel.ch. )
PHR H CHP,i
DE Power plant
CHP unit
H CHP,i DH
η boil
H boil,max LL boil.,i DH Boiler
η boil Auxiliary
boiler H boil,max LL boil.,i

Fig. 1. Redundant CHCP superstructure.

to have installed a CHP unit with 450 kWe and 600 kWh capacity PHRCHP DH;max
PESi ¼ CONStot;i   HCHP;i  LLboil;i  Ael:ch  LLel:ch:;i
(ge,CHP = 0.33) and an absorption chiller with 250 kWc capacity, g gboil
we have ð22Þ
250 where
HCHP ¼ þ 80 ¼ 437:1 kWh
0:7
) PHRCHP  HCHP ¼ 291:4 kWe ; C abs ¼ 250 kWc ð14:a; bÞ 1 1 1 DC;max
¼  ; ¼ Ael:ch ;
DC;max  LLel:ch: 60 g gE;CHP gpow:plant COPel:ch: gpow:plant
¼ ¼ 24 kWe DE
COPel:ch: 2:5 CONStot ¼ CONSconv  ð23:a; b; cÞ
60 gpow:plant
DEexch: ¼ 291:4  70  ¼ 197:4 kWe ð15:a; bÞ
2:5
  The parameters in Eq. (23) are calculated once only, before
80 1 310
CONSconv ¼ þ  70 þ ¼ 529:8 kWprim ; starting the optimization routine.
0:9 0:44 2:5
CHP 291:4 2.2.2. Basic constraints
Fuel ¼ ¼ 883:0 ð16:a; bÞ
0:33   The definition of an appropriate set of constraints requires for-
1 1 mulating energy balances for each plant component and express-
PES ¼ 529:8  291:4  
0:33 0:44 ing their technical limits in operation; boundary conditions
70 60 should be also expressed. All these constraints are here indicated
  ¼ 95:5 kWprim ð17Þ as ‘‘basic”, because they have an intuitive nature and directly de-
0:44 2:5  0:44
conv DEexch: rive from a correct analytical representation of reality. In Section
Fuel ¼ CONSconv þ 3, a different kind of constraints will be described, derived from
0:44
197:4 accurate energetic analyses and oriented to superimpose to the
¼ 529:8 þ ¼ 978:5 kWprim ð18Þ optimization routine some pre-determined results.
0:44
Fuel
conv
 Fuel
CHP
PES Covering energy loads
PESrel
% ¼ conv ¼ conv
Fuel Fuel C abs;i þ LLel:ch:;i  C el:ch:;nom ¼ Dc;i ;
978:5  883:0 95:5 C abs;i
¼ ¼ ¼ 0:097 ð19Þ HCHP;i  þ LLboil;i Hboilmax þ Q TES;i P DH;i ð24:a; bÞ
978:5 978:5 COPAbs
conv CHP
Fuel  Fuel PES
PESabs
% ¼ ¼ Production limits
CONStrad CONStrad
978:5  883:0 95:5 PHRCHP  HCHP;i 6 ECHP;nom ; C abs 6 C abs;nom ;
¼ ¼ ¼ 0:180 ð20Þ
916:2 529:8 STORTES;i
6 V TES ; STORTES;i P 0 ð25:a; b; c; dÞ
Evidently, PESrel abs
% and PES% can significantly differ, being the lat-
qwater cp ðT del  T ret Þ
ter higher when surplus power production (i.e. positive DEexch) is
Operation/synthesis congruence
achieved. Also, PESrel
% has an upper technical limit in the value:
HCHP;i 6 jc  dCHP ; C Abs;i 6 jc  dAbs ; j HTES;i j6 jc  dTES
1
PESrel
% ¼ 1
h
gH;CHP gE;CHP

gH;CHP COPabs
 i ð26:a; b; cÞ
1
max gboil þg ; COPel:ch:
þ gE;CHP  g
pow:plant pow:plant
where the congruence constant jc is assigned a conveniently large
ð21Þ
value (in the order of 105, if HCHP, CAbs and HTES are in kW). Other
while PESabs
% might exceed this value. EABOT assumes the PEStotal
limits in the operation of plant components, expressed by Eqs. (2)
(sum of PESi, expressed in kWhprim, throughout a year), and not and (3), were threaten as discussed in Section 2.1. The energy bal-
the PESrel
% , as objective function in the energy saving-oriented opti-
ance of the heat storage presented in Eq. (8) is also included. As con-
mization. In order to reduce the calculations, Eq. (13) was rewritten cern the TES, another aspect should be underlined [13]: referring
as follows: the optimization to a standard year which is supposed to repeat
 Author's personal copy

A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016 3011

identically throughout plant life, the minimum level of charge of the The condition expressed by Eq. (28) leads to different trade-off
storage capacity must be imposed for the first and the last hour of conditions depending on the ratio:
the year:
Hdir
CHP
A¼ ð30Þ
STORTES;1 ¼ STORTES;Nhours ¼ STORmin
TES ð27Þ C abs
which evidently depends on the design variables ECHP,nom and
If Eq. (27) was not included, an unreasonably large VTES value would
Cabs,nom; being our analysis oriented to save computational re-
have been achieved (especially for a small number of days, i.e. small
sources, i.e. to superimpose a priori the optimal management strat-
‘‘Nhours”), because the optimization routine would have assumed to
egy (not related to the decision variables), it seems more convenient
consume as much ‘‘zero-cost” energy (stored at the begin of the
to compare two cases, the ‘‘combined heat and power” and the
year) as possible.
‘‘combined power and cooling” (Eqs. (31) and (32), respectively):

3. Advances in LP modelling by energetic analysis Hdir

Profit ¼ MPel  ðPHRCHP  Hdir boil
CHP Þ þ MPfuel 
CHP
gboil
In this section the possibility to formulate additional constraints dir
CHP PHR CHP  HCHP
derived from energetic analyses is discussed; in particular, a ther-  MPfuel P0 ð31Þ
ge;CHP
moeconomic analysis is performed, in order to detect a priori  
trade-off profit conditions to be used for the optimization of plant C abs C abs
Profit ¼ MPel  PHRCHP  þ
operation. Thermoeconomics of CHCP systems has primarily been COPabs COPel:ch:
used for cost accounting [3]: assuming a rational basis such as ‘‘exer- PHRCHP C abs
 MPCHP
fuel  P0 ð32Þ
gy destruction” for assigning a cost to each flow has allowed to ge;CHP COPabs
determine a ‘‘rational value” for each product of polygeneration sys-
tems. Here, on the contrary, a marginal cost analysis is proposed Two indicators, the Total Spark Spread cooling and heating,
without cost allocation, having been assumed impossible to vary (TSSc and TSSh), are derived (see Eq. (33)) from Eqs. (31), (32);
the PHRCHP, i.e. to privilege the production of the products ensuring the lower heating value of fuel is introduced in the TSS expressions
a higher unit profit. Let us refer to the simplified scheme in Fig. 2 (the to keep into account that MPfuel is usually expressed in [c€/Nm3] or
effects of TES’ exclusion in the analysis will be discussed later), [c€/kg] (for gaseous and liquid fuels, respectively) and MPE in [c€/
where HCHP is assumed as main operation variable. A low disaggre- kWh]:
 
gation level is adopted, treating the ‘‘CHP unit + absorption chiller” 3600PHRCHP
 MPboil PHRCHP
gboil HLVboil fuel þ MPE COPabs
þ COP1  MPE
system as a basic unit to write the cost balance; evidently, the ther- TSSh ¼ fuel
; TSSc ¼
el:ch:

mal output HCHP is represented as sum of three contributes: the heat

3600
ge;CHP HLVCHP
 MPCHP
fuel
3600PHRCHP
ge;CHP HLVCHP COPabs
 MPCHP
fuel
fuel fuel
covering thermal loads, the heat feeding the absorption chiller and,
ð33:a; bÞ
eventually (in case of non-heat-tracking operation), some surplus
heat rejected. In order to favour a correct interpretation of the above expres-
Trade-off profit conditions emerge by a comparison between sions, TSSh and TSSc are here calculated, as an example, by assum-
the operating cost in the separate and combined power production ing gE,CHP = 0.33, COPabs = 0.7, COPel.ch. = 2.5, MPE = 12 c€/kWh,
cases: MPfuelCHP = 0.35 c€/Nm3, MPfuelboil = 0.48 c€/Nm3:
 
C abs Hdir 36000:75
 0:48 þ 0:12  0:75
Profit ¼ MPE  PHRCHP  HCHP þ þ MPboil
fuel 
CHP
TSSh ¼ 0:9037000 ¼ 1:25;
COPel:ch: gboil 3600
0:3337000
 0:35
PHRCHP  HCHP 0:75 1
þ 2:5  0:12
 MPCHP
fuel P0 ð28Þ TSSc ¼ 0:7
¼ 1:71 ð34:a; bÞ
ge;CHP 36000:75
 0:35
0:33370000:7

Being a priori unknown Hwaste

CHP , the problem cannot be solved Evidently, at the examined conditions, combined heat and
yet. However, from the fundamentals of thermoeconomics we power have lower margins for profit with respect to combined
know that cost should be only allocated to useful products; conse- power and cooling. We may conclude that whenever, throughout
quently, two cases can be distinguished, respectively, when some a year, TSSh (the lower of the two TSS) is higher than 1, operating
surplus heat is recovered and rejected and when not: the CHCP plant in combined production mode (at any value of the
Hdir
CHP =C abs ratio) is profitable; of course, this result is valid as far as
– If no surplus heat: no surplus heat is produced.
C abs
Hwaste
CHP ¼ 0 HCHP ¼ Hdir
CHP þ ð29Þ – If surplus heat (Hwaste
CHP 6¼ 0):
COPabs

Fig. 2. Zoom on energy flows involving the examined section of the CHCP scheme.
 Author's personal copy

3012 A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016

When the CHP unit operates in ‘‘power production mode”, i.e. large TES are excluded, being most of their stored energy con-
the heat recovery is not used to cover any useful demand, profit sumed in off-peak hours when both SS and TSS are lower than
must be expressed considering the only useful output, as follows: 1; an ex-post analysis will be used validate this hypothesis.

PHRCHP  HCHP
Profit ¼ MPel  PHRCHP  HCHP  MPCHP
fuel P0 ð35Þ Before applying the method to real buildings, let us briefly dis-
ge;CHP cuss what effects additional constraints are expected to have on
Trade-off conditions can clearly expressed with reference to the the optimization algorithm. With reference to Fig. 3a and b, where
Spark-Spread: representations of linear and non-linear two-variables problems
are given, it is evident that in LP problems the optimal solutions
MPE lie on the boundary of a feasible-region polytope (n-dimensional
SS ¼ ð36Þ
3600
ge;CHP HLVCHP
 MPCHP
fuel space enclosed by a finite number of hyperplanes), i.e. at the inter-
fuel
section of the binding (active) constraints; on the contrary, in non-
Again, SS values higher than 1 indicate the profitability of auto- linear programming (NLP), solutions from both the boundary and
producing electricity, with respect to the option ‘‘switch the engine within the feasible-region are accommodated. Consequently, large
off and buy electricity”. benefits could be expected when NLP algorithms are implemented,
Obviously, Eq. (35) is a much more restrictive condition with re- due to the reduction of the feasible-region polytope; on the con-
spect to Eqs. (31), (32); consequently, in many countries the former trary, as concerns LP techniques the benefits could either be rele-
condition is expected to be fulfilled in peak-hours only, while Eqs. vant or negligible, depending on the search direction adopted. In
(31) and (32) can be also fulfilled in medium and, eventually, off- particular, if an interior point method is adopted (like in our case),
peak hours. The above described method is very easy to imple- some reduction in computational resources could be achieved
ment; any optimization tool like EABOT can be equipped with a while no benefits are expected if the simplex method is used [21].
preliminary routine which calculates SS, TSSh and TSSc for any hour
of the year and can thus perform a ‘‘conditioned optimization”, 4. Applying EABOT to two cases-study
with a pre-fixed plant operation. Two aspects should be here
underlined: Optimizations were performed by EABOT for a 646 rooms (1146
beds) hotel and a 300 beds hospital situated in Italy, whose elec-
1. The auxiliary production units were neither included in Fig. 2 tricity, heating and cooling loads have been presented in [6,22];
nor in the proposed analysis. This is why zero profit is associ- the measured energy load peaks are presented in Table 1. As con-
ated with these units and their influence on the trade-off profit cerns energy prices, two different scenarios are considered, both
conditions is null. assuming a four-bands electricity price structure (peak, high load,
2. We should ask what changes in the above analysis when we medium load and off-peak hours); the first scenario takes into ac-
take into account the presence of the TES. Evidently, when SS count the tax exemption for CHP fuel, the second does not. The val-
is higher than 1 nothing changes: operating the CHP unit at full ues adopted for the optimization, referring to the Italian case, are
load remain the most profitable choice. Similarly, when TSS is resumed in Table 2.
lower than 1, the CHP unit should be switched off even in pres- The EABOT tool was systematically applied to the two examined
ence of the TES. The only deviation from the above described buildings, assuming different values for Nhours (in any case, multi-
analysis concerns the hours where TSS is higher than 1 and SS ples of 24, being the analysis performed on hourly basis); five dif-
lower than 1: in fact, the surplus energy eventually stored dur- ferent conditions were tested, that are
ing peak hours (SS > 1) could be sufficient to cover the heat
loads in other hours with SS lower than 1 and TSS higher than
1. While in absence of the TES the heat recovery could have
been covering a useful demand (requiring to use the TSS expres- Table 1
sion), in presence of the TES the SS should be more properly Measured energy load peaks of the two examined buildings
adopted to assess the most rational operation. This analysis
Electricity (kW) Heating (kW) Cooling (kW)
slightly complicates the problem; here, for sake of simplicity,
Hotel 985 3,134 1,129
the effects of the TES on the proposed operation criteria will
Hospital 241 2,655 1,548
be neglected. This approach is some way reasonable if very

Fig. 3. Search directions for a 2-variables optimization in case of (a) linear programming, (b) non-linear programming.
 Author's personal copy

A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016 3013

Table 2 (c) Profit-oriented optimization with tax exemption for CHP

Energy prices assumed for the optimization fuel and an additional constraint ensuring the solution to
Prices SS and TSS SS and TSS achieve a minimum PESrel % (as required by the Directive
(if tax exemption) (if no tax exemption) 2004/8/EC).
Electricity (d) Primary energy saving-oriented optimization (max PESabs% ,
Peak hours 0.16 (€/kWhe) 1.27 and 1.89 1.02 and 1.51 regardless of economic viability).
High load hours 0.105 (€/kWhe) 0.84 and 1.45 0.67 and 1.16 (e) Profit-oriented optimization with tax exemption and a pre-
Medium load 0.07 (€/kWhe) 0.56 and 1.17 0.45 and 0.94
hours
fixed operation strategy (see Section 3) superimposed.
Off-peak hours 0.045 (€/kWhe) 0.36 and 0.97 0.29 and 0.78
Natural gas
The results obtained in terms of ECHP,nom, Cabs,nom, VTES, NPV and
No tax exemption 0.50 (€/Nm ) 3 annual PESabs
% are plotted in Fig. 4 versus the number of days used
With gas 0.40 (€/Nm3) for the optimization; in Fig. 5 the calculation time, referred to an
exemption ordinary performance machine (3 GHz processor, 1024 GB RAM),
is also represented. Let us observe that

– All the main synthesis and design variables converge to

(a) Profit-oriented optimization, with tax exemption for CHP quite constant solutions when the number of days used for
fuel. the optimization is higher than 15–20.
(b) Profit-oriented optimization, with no tax exemption for CHP – The consumption of resources (i.e. the calculation time) rap-
fuel. idly increases when the number of days is higher than 30;

Fig. 4. Optimal ECHP, Cabs,nom,VTES, NPV and PESabs

% versus the number of days used as a basis for the optimization.
 Author's personal copy

3014 A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016

The comparative analysis revealed that a number of days rang-

ing between 24 and 30 already ensures a high reliability of results;
then, the results shown in the following of this paper will be re-
ferred to the assumption of 24 days (Nhours equal to 576). As con-
cerns the energetic results:

– Comparing the results of the options ‘‘a” and ‘‘b” (the option
‘‘c” is not represented in the figure), the relevance of tax
exemption to make trigeneration viable is evident. Very
small CHP units resulted in the no incentives case, leading
to exclude the absorption chiller from the plant lay-out
and consequently to achieve quite small energy and money
savings.
– The PES-oriented optimization (option ‘‘d”) leads to oversize
the combined production unit and the TES, as could be
Fig. 5. Time consumed for the optimization as a function of the number of days.
expected. In fact, being any economic index excluded from
the decision function, a convergence towards negative NPVs
even considering 60 days, however, the calculation time is may be observed. As concerns the PESabs % , high values are
quite small due to the relevant simplifications introduced achieved (in the optimization of the CHCP plant for the hos-
in the model.

Fig. 6. Optimal operation for three consecutive winter days and dependence on energy prices.
 Author's personal copy

A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016 3015

pital, not shown in the figure, values higher than 0.35 also In order to put into evidence the rational of the results obtained,
resulted!), which can be explained by the analysis presented a zooming representation of results is preferable to a full scale one.
in the Section 2.2.1. Consequently, in Figs. 6 and 7 the results are shown for three con-
– As concerns all the 5 monitored parameters, the optimiza- secutive days in the winter and four consecutive days in the sum-
tion options ‘‘a” and ‘‘e” converged to the same solutions; mer period, respectively; representing the results for all the
hence, at synthesis-design level the superimposed operation optimized 24 days would lead to a non-clear plot, while zooming
strategy does not affect the reliability of results (with the the results for one single days would make us loose any possibility
additional constraint, a convergence to the same ‘‘optimal to detect the mutual influence of TES charge between two consec-
solution” as in the ‘‘no additional constraint” case can be utive days.
observed in a lower computation time).

Fig. 7. Optimal operation for four consecutive summer days and dependence on energy prices.
 Author's personal copy

3016 A. Piacentino, F. Cardona / Energy Conversion and Management 49 (2008) 3006–3016

Both figures, referring to the optimization option ‘‘a” (profit ori- strategy. Implementing the proposed approaches into a tool al-
ented with incentives), allow us to recognize that the spontaneous lowed us to perform multi-objective optimizations for two large
convergence of the optimization reflects the operation strategy buildings in the civil sector and to derive a few conclusions about
superimposed in the option ‘‘e”. In Fig. 6, the CHP unit is operated the optimal number of days to be used for the optimization (which
at full load when TSS is higher than 1 and switched off in other resulted from a compromise between the objectives of results’ reli-
cases. However, this occurs just because a useful heat demand ex- ability and fastness of the optimization) and the consistence of the
ist, as evident when looking at the HTES curve, which shows that hypotheses introduced. This paper was only intended to offer a
any heat excess is used to charge the tank; in fact, being SS lower new perspective on the problem of the improvement of MILP tech-
than 1 (no peak, just high load hours are present in winter days), niques for CHCP optimization; such analysis must be evidently
operating the unit at an as high load level as to reject surplus heat coupled with those most typically proposed by operational re-
would not be convenient. Evidently, even including the TES, super- search specialists.
imposing the operation strategy in peak (SS > 1, engine at full load
operation) and off peak (TSS < 1, engine switched off) hours allows References
the LP algorithm to converge to the optimal solution, being vali-
dated the hypothesis made in Section 3 to neglect TES’ influence. [1] Hernández-Santoyo J, Sánchez-Cifuentes A. Trigeneration: an alternative for
Also, looking at the STORTES we may observe that the presence of energy savings. Appl Energy 2003;76:219–27.
[2] Ziher D, Poredos A. Economics of a trigeneration system in a hospital. Appl
long peak-hours periods leads to very high peaks of the stored en- Therm Eng 2006;26:680–7.
ergy; in Fig. 6, almost 8,000 kWh, 5 times more than in the summer [3] Temir G, Bilge D. Thermoeconomic analysis of a trigeneration system. Appl
period (see Fig. 7) where the energy price profile is more irregular. Therm Eng 2004;24:2689–99.
[4] Teopa Calva E, Picón Núñez M, Rodrı́guez Toral MA. Thermal integration of
In Fig. 7 the cooling balance is also plotted; it is evidently influ- trigeneration systems. Appl Therm Eng 2005;25:973–84.
enced by the heat recovery rate HCHP, which ultimately depends [5] Dimopoulos GG, Kougioufas AV, Frangopoulos CA. Synthesis, design and
on the energy price. During peak hours the CHP unit operates at operation of a marine energy system. Energy. doi:10.1016/
j.energy.2007.09.004.
full load regardless of the existence of a useful heat demand; how- [6] Cardona E, Piacentino A. A methodology for sizing a trigeneration plant in
ever, again no heat is rejected due to the sufficient capacity VTES of Mediterranean areas. Appl Therm Eng 2003;23:1665–80.
the storage tank. Let us observe that a note in the PESabs
% curve (see
[7] Cardona E, Piacentino A. Optimal design of CHCP plants in the civil sector by
thermoeconomics. Appl Energy 2007;84:729–48.
Fig. 7) indicates the scarce energetic relevance of its singular
[8] Kong XQ, Wang RZ, Huang XH. Energy optimization model for a CCHP system
points; in fact, being it artificially PESabs
% calculated on hourly basis, with available gas turbines. Appl Therm Eng 2005;25:377–91.
depending on whether the TES is charging or discharging, very high [9] Oh SD, Lee HJ, Jung JY, Kwak HY. Optimal planning and economic evaluation of
(positive or negative) energy saving rates could result, which have cogeneration systems. Energy 2007;32:760–71.
[10] Gamou S, Yokoyama R, Ito K. Optimal unit sizing of cogeneration systems in
no relation with the actual energy saving potential in the examined consideration of uncertain energy demands as continuous random variables.
hour. Actually, a detailed analysis which cannot be presented here Energy Convers Manage 2002;43:1349–61.
would detect minimum differences in the optimal operation [11] Rong A, Lahdelma R. An efficient linear programming model and optimization
algorithm for trigeneration. Appl Energy 2005;82:40–63.
achieved according to the options indicated as ‘‘a” and ‘‘e”; how- [12] Rong A, Lahdelma R, Luh PB. Lagrangian relaxation based algorithm for
ever, the reduction in the consumption of resources can be signif- trigeneration planning with storages. Eur J Oper Res. doi:10.1016/
icant (depending on the algorithm adopted) and it absolutely j.ejor.2007.04.008.
[13] Dotzauer E, Holmström K, Ravn HF. Optimal unit commitment and economic
justifies the use of additional constraints. The discussion of results dispatch of cogeneration systems with a storage. In: Proceedings of the 13th
(Figs. 4, 6 and 7) presented in this subsection ensures the system to PSCC conference, Trondheim, Norway; 1999.
have converged to ‘‘energetically meaningful” solutions; the expe- [14] Piacentino A. CHP and CHCP applications in buildings: energetic, exergetic and
economic analysis of the possible solutions. PhD thesis, Palermo, January 2005
rience of the designer is decisive in this phase of code testing. [in Italian].
[15] Arcuri P, Florio G, Fragiacomo P. A mixed integer programming model for
5. Conclusions optimal design of trigeneration in a hospital complex. Energy 2007;32:
1430–47.
[16] Piacentino A, Cardona F. Integrated optimization of synthesis, design and
Starting from the assumption that large progresses have been operation in CHCP-based l-grids – Part I. Description of the method.
made in the use of efficient algorithms for the optimization of syn- In: Proceedings of ECOS 2007, Padova. Italy: SGE Pub.; June 2007. p.
575–84.
thesis, design and operation of a CHCP system including thermal [17] Cardona E, Piacentino A. DABASI-WWW promotion of energy saving by CHCP
energy storages, a new approach was proposed in this paper, focus- plants – database and evaluation. SAVE II Project. Contract No. 4.1031/Z/02-
ing the attention on the energetic analysis of the plant. Considering 060, Bruxelles; 2005.
[18] Valero A, Serra L, Lozano MA. Structural theory of thermoeconomics. In: ASME
that MILP optimization of trigeneration systems represents a quite book-AES, Thermodynamics and the design, analysis, and improvement of
approximate approach due to the linearization of components’ energy systems, vol. 30; 1993. p. 189–98.
behaviour and cost figures, some simplifications were introduced, [19] Kostowski W, Skorek J. Thermodynamic and economic analysis of heat storage
application in co-generation systems. Int J Energy Res 2005;25:177–88.
like the exclusion of binary variables for the hour by hour unit
[20] De Paepe M, Mertens D. Combined heat and power in a liberalised energy
commitment problem which significantly reduces the consump- market. Energy Convers Manage 2007;48:2542–55.
tion of computational resources. An in-depth analysis of trade-off [21] Venkataraman P. Applied optimization with MATLAB programming. New
York: Wiley; 2002.
profit conditions was proposed, partially based on thermoeconom-
[22] Piacentino A, Cardona F. On thermoeconomics of energy systems at variable
ics, allowing us to formulate an artificial (i.e. non-physical) con- load conditions: integrated optimization of plant design and operation. Energy
straint which essentially superimposes a pre-fixed operation Convers Manage 2007;48:2341–55.