Thermoptim: from learning to exergy optimization and

systemic of combined cycles and cogeneration installations

and trigeneration based on gas turbines
R. Gicquel
School of Mines of Paris, 60 Boulevard Saint Michel
75272PARIS CEDEX 06, France
[Link]@[Link]

This document briefly presents some possible uses of the software Thermoptim for the study of
gas turbines and associated cycles. It is structured in two parts followed by an appendix:

- The general introduction summarizes the main features and applications of Thermoptim.
- The second part specifically addresses cycles involving gas turbines.
- An appendix provides a more detailed presentation of the software package.

1 General introduction
The software Thermoptim ([Link])
is primarily a modeling environment
systematic energy technologies very
original and without equivalent, which makes possible
the adoption of new methods particularly
fertile in terms of pedagogy and modeling
and optimization.

Figure 1: Schematic of a steam cycle

The same tool is now being implemented by around sixty users for two main categories of applications:
either educational or industrial.

By assembling predefined or specifically developed component models, it allows for a very accurate representation.
easily very varied energy systems (figure 1), from the simplest to the most complex. Depending on the cases, the models
These components can be either purely phenomenological or more technological, that is to say capable
to carry out sizing or to simulate operation under non-nominal conditions. Thermoptim stands out
other component-oriented software due to the emphasis placed on the coupling and interactions between these components, it is-
to discuss the systemic properties of the studied installations. It is thus complementary to the specialized tools developed.
by the manufacturers for the development and manufacturing of their devices.

The libraries and kernel computation methods enable it to extensively cover the field of energy.
(steam cycles, refrigeration cycles, combustions, wet mixtures ...), and an expansion mechanism gives it
great possibilities in terms of interoperability: addition of external components, customized control of its calculations... More
Details on the features of Thermoptim are provided in Appendix 1.
 2

1.1 Educational applications

Used in more than 50 higher education institutions since 1997, Thermoptim has proven to improve
significantly the teaching of energetics, in a way to do thermodynamics without difficulty.

Considered until now to be particularly difficult by generations of students and engineers, thermodynamics
applied to energy systems now has an original and extremely simplified learning method
based on the use of Thermoptim, in which, to study energy technologies, the emphasis is placed more on
on the system and qualitative aspects as well as on the detailed analysis of component behavior and quantification of
their performances [1, 2, 3, 4].

Research in the didactics of applied sciences has shown that the teaching of a discipline is rooted in a
double reference corresponding on the one hand to certified scientific knowledge as produced by the
research laboratories (we often talk about knowledge or declarative knowledge), and on the other hand to knowledge
related to techniques and their applications in society (we then talk about know-how or knowledge
procedural) [5]. The former are rather abstract and pertain to theory, while the latter are more practical and
action-oriented. To succeed in learning a discipline, it is increasingly accepted that the
students must build their knowledge based on their previous acquired skills, through a back-and-forth between these two
standards, modeling being one of the preferred means to do so.

For energy systems, it is unfortunately almost always impossible to find models that are both simple
and precise. By barely caricaturing, one could say that the classical approaches to the discipline are faced with a
dilemma, the models they lead to being either unrealistic or incalculable. This explains why the
Thermodynamics is often considered a difficult subject to teach. The problem has been identified for a long time.
and many efforts have been made to address this, but until recently there were still a lack of solutions, and this
malgré les efforts déployés par les enseignants et les évolutions des programmes.

Thermoptim allows to represent many energy technologies by simple graphical assembly of

predefined components. Without writing a single line of code, it is possible to establish phenomenological models for
calculate their performances with very high precision and plot their cycles in the various diagrams
thermodynamics. Freed from the usual computational difficulties in the subject, students are much more motivated.
for discipline and they assimilate much better.

Implemented for several years in the various training cycles of the Mines Paris School, and used at the end
2002 in about fifty higher education institutions (preparatory classes for grandes écoles, IUT,
Universities and engineering schools, in France and abroad, this pedagogy complements the classical approach in
inviting students to start studying energy systems by exploring or developing them themselves
models of the main energy conversion technologies, thanks to the software Thermoptim. They thus have
a platform for virtual experiments allowing them to develop models of numerous energy systems
by going beyond the limits we have just mentioned.

Based on the experience gained in recent years, we can group the approaches into three main categories.
possible pedagogical approaches when one has a tool like Thermoptim:
- guided exploration activities, in which the student simply views pre-assembled models in a
catalog of examples,
- the activities of assembling models from predefined components, in which he must build himself and
set up the system that he studies,
- and finally the addition of new components to those of the existing core, which requires him to write the model himself
of the component (selection of characteristic parameters, coupling variables, and the appropriate set of equations,
definition of the graphical interface and computer implementation.

The first category is aimed at beginners, especially if they have little time and are working alone, for example.
by following an online training module. The second requires more time, as well as a deeper learning of
Thermoptim, particularly from the management, which must be able to correct construction errors of the models. The
third finally addresses students already familiar with the software and sufficiently proficient in thermodynamics to
to be able to start designing component models. They also need to have a minimum of knowledge in
programming with Java or at least C. Interfacing with Fortran is also possible.
 3

1.2 Industrial applications

Increasingly, energy technologies are appearing as complex interconnected systems with functions
multiple, these systems themselves being very deeply integrated and having to adapt to evolving specifications
during their lifespan. To study them, it is desirable to have environments that facilitate and
to secure their modeling as well as to optimize their internal structure.

Combining a systemic approach with a classic analytical and/or empirical approach, Thermoptim offers a potential
significant in this context. Its graphic environment allows for the visual construction of models of many
energy systems, from the simplest like a refrigerator to the most complex like power plants
combined cycle with integrated coal gasification involving several hundred elements.

At the industrial level, Thermoptim facilitates and secures the study of energy systems by simplifying the approach to
modeling and making it safer, and its exerghetic method is used in about ten companies like
Framatome, EDF or CEA to optimize complex systems.

1.2.1 Facilitate and secure the modeling of energy systems

Thermoptim allows its users to easily calculate even complex thermodynamic cycles, and this
without having to write a single equation or program if they use the predefined components in the software package core.

Not only does such a way of doing simplify the modeling process significantly, but it also facilitates it later on.
the use and maintenance of the model, but above all it secures its construction by automating the establishment of
couplings between the different elements that make it up and ensuring their coherence. This last point is essential
for large systems involving several hundred or thousands of equations whose verification needs to be
automated in one way or another.

Based on the experience accumulated by its users and the expectations they have expressed, it has become desirable
to extend the capabilities of the tool so that it can solve problems even more difficult than those for which it
was initially designed: the advanced modeling of certain energy systems exceeded the capabilities of the versions
of Thermoptim disseminated until the end of 2002, the limits concerning mainly two dimensions:
on the one hand, the scope covered by the software package: only the predefined components available on the editor's palette
diagrams could be brought into play in the studied energy systems. Even if they allowed to represent the
Most of the energy technologies used in practice could not be modeled in Thermoptim.
In the same spirit, some users had expressed the wish to be able to modify the available templates in
the tool by introducing their own equations, which was until recently impossible;
On the other hand, the finesse of modeling: the constructed phenomenological models certainly allowed for the study of the
thermodynamic cycle of the studied technology, but not to carry out a precise technological sizing, nor
to simulate the performance in non-nominal operating conditions, these last two issues being
much more complex than the first.

New features have been incorporated into the software to allow it to exceed these limits. To broaden the
the domain covered by the software, we now distinguish on the one hand the internal core of Thermoptim, which includes all the
features of previous versions, and on the other hand an extension mechanism that allows a user to add to
core of custom elements (new bodies, new components…). Users can thus customize
their tool by defining bodies or components not available in the Thermoptim core, which integrate
perfectly in the environment: it becomes possible to study the integration of components into complex systems
like fuel cells, chemical reactors, absorption machines... Furthermore, in order to refine the
modeling, new screens, called technological sizing, as well as algorithms were introduced
calculation in non-nominal regime.

1.2.2 Optimiser les systèmes énergétiques complexes grâce aux méthodes exergétiques

Classic approaches to energy technology optimization, applicable for optimizing one by one the
Different components of a system are insufficient to guide the designer in choosing the best one.
complete system configuration. The optimization of internal thermal couplings through pinch analysis.
 4

"analysis" provides a rigorous method for this, allowing for the optimization of the overall configuration of the installation, and
by ensuring the best consistency between all energy needs and availability.

Systemic integration optimization, which we have been working on for about ten years, falls under the methods
exergies. As soon as we assemble various components together to form a system, irreversibilities
particular ones appear in addition to those of each element. These irreversibilities can be qualified as systemic because
they mainly depend on the relative positioning of the components. Reducing these irreversibilities allows
to increase the efficiency of the systems considered, mainly through better internal regeneration. This work,
presented in more detail in the last part of this note, have led to several publications and to
collaborations with industrial partners (Babcock, EDF, Framatome). They enabled the implementation in Thermoptim of a
set of well-validated algorithms.

2 Illustrations of the use of Thermoptim for cycle study

gas turbines
In this section, we will start by indicating how Thermoptim can be concretely used to show to some
beginner students how to realistically model a gas turbine [6], disregarding the assumption of gases
perfect, then we will provide some suggestions for related educational activities.

Then, we will present a number of cycle models involving gas turbines and created with the help of
of Thermoptim. Due to lack of space, and the main objective being to illustrate the possibilities of the software package, these
models will only be documented very superficially, external references being provided where they are
available. We will successively present:
- a combined cycle at a pressure level
- an industrial cogeneration installation
- a trigeneration installation
- a gas turbine associated with a SOFC type fuel cell
- the coupling in Thermoptim to an IGCC model of CO conversion units and CO capture 2

Finally, we will discuss the implementation of the optimization method by presenting a combined cycle at two levels of
pression.

2.1 Simple gas turbine model (learning)

2.1.1 Development of the model

Figure 2: Qualitative graphical description of a gas turbine

A simple gas turbine consists of three interconnected elements: an air compressor, a combustion chamber, and a
turbine. To represent them, simply select these three components from the palette of the schematic editor, and then
place on the work screen. Moreover, to be able to properly represent things, it is preferable to plan for two arrivals of
fluids, one for air and the other for fuel, and an outlet for the burnt gases.

In the first step, these six components are successively placed on the screen. During the placement, a
A minimal description of each component is requested from the user: name of the component and the output point, name of the
fluid and flow passing through the component.
 5

In a second step, the connections between these components are established by clicking on the output port of one of them.
and by sliding the cursor to the input port of the downstream element. During this connection, various information is
propagated from upstream to downstream. We then obtain the diagram of figure 2, which constitutes the qualitative model of the gas turbine.
From a didactic perspective, such a qualitative representation is essential for properly assimilating the principles of design.
technologies.

In a third phase, a
interface allows you to create from
this diagram is the set of points and
transformations corresponding to
problem described, and, in duplicate-
by clicking on each component, it is
possible to quantify the model in
parameterizing each of these elements
to refine the description then
perform the calculations.

Figure 3: Compression calculation screen

Figure 3 shows, for example, the screen related to compression. It should specify whether the transformation occurs in the system.
open or closed, if it is represented by a polytropic law (like here) or an isentropic efficiency, enter the value
the corresponding yield...

Figure 4: Schematic of the gas turbine

Finally, once all the calculations are done, the performance of the cycle is determined and the state of the different points can be
Displayed in the diagram (figure 4). Depending on the combustion settings, the composition of the burnt gases varies.

Once this model is created, sensitivity analyses can be conducted, for example to study the influence of the ratio of
compression on the machine's performance.

Figure 5: Gas turbine cycle on diagram (T,s)

 6

The representation of this cycle on the different thermodynamic diagrams poses, however, a small interesting problem.
to discuss with the students, due to the change in the composition of the working fluid in the combustion chamber: on the
enthalpy diagram of air (figure 5), points 3 and 4 do not appear on the correct isobars, due to the fact that their state has
was calculated with the thermodynamic properties of the burnt gases and not those of air. Strictly speaking, one should not
to represent such a cycle on the diagram of a single gas, although it is done frequently.

2.1.2 Energy and exergy balance

Based on the open cycle gas turbine model developed under Thermoptim, it is easy to draw up the energy balance of
Cycle. The efficiency here is equal to the ratio of the work performed by the cycle to the heat provided by combustion.

To establish the exergetic balance, one must start by setting a reference temperature and pressure. We
We will take here
0 T = 288.15 K (15 °C) and
0 P = 1 bar, corresponding to the air intake conditions in the gas turbine.
The reference temperature having entered the Thermoptim configuration screen, the cycle must be recalculated, and the
exported results to be easily exploited.

The exergy balance of the combustion chamber is easily established: per kg of intake air, the fuel releases an exergy
equal to 890 kJ. The variation of exergy of the fluids in the chamber being equal to 614 kJ, the irreversibilities are equal to
276 kJ. The exergonic efficiency of the chamber is therefore equal to 69%, while its energy efficiency is 100%.
because it is adiabatic. If we had taken into account pressure losses, this efficiency would be slightly lower.
Finally, since the cycle is not closed, one must take into account the residual exergy of the exhaust gases, which adds to the
losses due to irreversibilities.

Figure 6: Energy and exergy balances of a simple gas turbine

In figure 6, the losses are given as a percentage of the total irreversibilities. It shows that these are
mainly concentrated in the combustion chamber and in the exhaust, and that they are substantially equal in
the compressor and the turbine (their polytropic efficiencies are equal).
The improvement areas for gas turbines are derived from this exergy balance:
increase the cycle temperature to enhance combustion efficiency. We discussed the limits
technological aspects that one encounters at this level;
recover the available exergy at the exhaust.
 7

2.1.3 Related educational activities

0.38
r = 27
0.37
0.36 r = 16

0.35
0.34
r = 11
0.33
0.32
0.31
250 270 290 310 330 350
net work (kW)

Figure 7

Around the gas turbine cycle, many activities that interest students can take place.
compare the efficiency of the cycle with that of the Carnot cycle operating between 15°C (cold source) and 1200°C
(maximum temperature of the cycle), to have an idea of the deviation of the real cycle (0.358) compared to the ideal cycle
(0.797)
study the evolution of the cycle's efficiency and the power produced when the inlet pressure and temperature change
turbine variant. By varying the cycle pressure, we obtain the abacus of figure 7 showing the evolution of
efficiency as a function of useful work. The maximum efficiency is achieved, at a given turbine inlet temperature, for
a compression ratio of 27, and the maximum power for 11;
Most gas turbines operate in an open cycle and release high-temperature gases (500) into the atmosphere.
- 600°C). It can sometimes be possible to improve the cycle by using regeneration, that is to say by
valuing the exhaust gases by preheating the air from the compressor outlet. A discussion on the limits of this
The solution provides an opportunity for a fruitful exchange with the students;
Another way to improve the cycle is to inject steam generated using exhaust gases to
vaporizing pressurized water in liquid state;
a model based on the assumptions of the common competition subject 2001 of the schools of mines of Albi, Alès, Douai and
Nantes (turbine gas crossed by supposed perfect air) can also be presented, along with the results to which it
conduit for some compression ratios (4, 8, and 12), and the representation of the corresponding cycles on a
entropic diagram equipped with isobars (figure 8). This model can be compared to the previous one, which is much more accurate.
due to the fact that combustion is indeed calculated.

Such studies of variants

can be proposed,
notably based on the
eight examples proposed in the
chapter 2 of reference [2], which
contain, in addition to a turbine at
regeneration gas, and a turbine
steam injection gas that
have just been mentioned:

Figure 8: Representation of gas turbine cycles with perfect air for three compression ratios
a helium closed cycle gas turbine, the interest of this example being that helium is a perfect gas, allowing
to achieve the same results with the software and with an analytical approach.
a gas turbine intermediate cooling
a sequential combustion gas turbine
a two-stage gas turbine with intermediate cooling, heating, and regeneration. This example, much more
 8
More complicated than the previous ones, it involves two exchangers (cross type, non-mixed), one of which creates a loop.
internal: the regenerator;
a refrigerated gas turbine

Figure 9

2.2 Combined cycle at a pressure level

The excellent yields that combined cycle power plants achieve today (over 60% on lower heating value),
are the result of the integration into a single production unit of two complementary technologies in terms of level
of temperature: gas turbines, which operate at high temperatures (in an aeroderivative machine the gases enter
typically at 1300 °C in the expansion turbine, and exit at around 500 °C), and steam power plants, which operate at some
lower temperatures (between 450 °C and 30 °C in this case).

Figure 10: Combined cycle at a pressure level

The principle of a combined cycle consists of operating one or more gas turbines in cascade, followed by a
steam power plant whose hot source is the cold source of the gas turbines (figure 9). Under these conditions, the gases
The exhaust from the gas turbine is used in a recovery boiler where steam is produced that is
then relaxed in a condensing turbine.

The combined cycle obtained is a particularly successful marriage in the pursuit of improving efficiency.
thermal: with the currently available equipment, the efficiencies achieved exceed 55% and are therefore higher than
those that can be hoped for, even in the medium term, from the most advanced future steam power plants.
 9

A simple yet interesting exercise for students is to determine the ratio of mass flow rates between the
two cycles to best cool the exhaust gases (figure 10).

We can thus show them that with a single pressure level cycle, it is impossible to fully cool the gases.
coming out of the turbine. The design of a two or three-pressure level cycle is greatly facilitated if one uses
the optimization method of Thermoptim will be discussed later.

The energy and exergetic balances of this combined cycle at a pressure level (provided in the catalog of examples of
Thermoptim) appear in figure 11.

2.3

Figure 11: Energy and exergonic balances of a combined cycle at a pressure level
Installation of industrial cogeneration

The cogeneration installation (existing) that we present here is relatively complex: it produces electricity and
provides heat to the district heating network of a city with 30,000 inhabitants. It is a "heat-" type installation.
force", where heat is the base product, electricity being a by-product resold to EDF allowing for better
value the fuel [7].

The heating network is a water circuit under

pressure with a temperature ranging from about
90 °C and 130 °C. The circuit starts from the power plant to its
maximum temperature, then he travels around the city and
feeds exchange substations themselves
intended to provide heating and/or reheating
sanitary water from residences, buildings, hospitals,
of school groups... Finally, he returns to the central at his
minimum temperature, to be warmed up.

Figure 12: Diagram of the cogeneration plant

The heat needs of the city vary throughout the year depending on climatic conditions. The plant must
adapt to these needs and adjust the thermal power it provides to the network.
 10

The installation consists of three independent circuits that exchange only heat with each other (figure 12).
A gas turbine (GT), connected to a generator, accounts for about 80% of the electricity production at the site. The gases that
escapes from the TAG at about 450°C are burned a second time thanks to a low fuel supply. After
the post-combustion, the fumes, which have a temperature of about 600°C, heat the water of the steam circuit in a
recovery boiler, then water from the urban network. Finally, they are released into the atmosphere at approximately 150°C.
through the chimney;
a steam circuit
ensure the remaining 20%
of the production
electric. The water is
heated, transformed into
vapor, then overheated
in the boiler of
recovery by the
exhaust fumes
from the TAG. A
desuperheating is
carried out in the middle of the
overheating in order to
regulate the temperature
from the steam at the entrance of
the turbine. It is
then relaxed in
a steam turbine
(TAV) connected à a
alternator. In the
condenser, crossed at the
times through the steam circuit
and by the urban network,
the steam condenses into
warming this one.
Finally, liquid water is
directed towards the tarpaulin
food, and pumped
towards the boiler of
recovery

Figure 13: Results of the cogeneration installation simulation

The urban network is traversed by hot liquid water under pressure. The latter returns to the plant.
minimum temperature. It is then heated in the condenser, where it acts as a cold source. Part of
this water is heated in the recovery boiler, in an exchanger called 'pin'. The flow rate in the pin
is limited. If the thermal inputs from the condenser and the pin do not meet the needs of the network, the supplement
is ensured by bypass boilers.

Figure 13 shows the model of the set of three circuits and their thermal interactions through the heat exchangers. The balance
energy of the plant, as well as the different efficiencies of the cogeneration installation: mechanical efficiencies (or
electric), thermal, global, and heat-power ratio, could be determined in different operating cases.
 11

2.4 Installation de trigénération à micro-turbine

A micro gas turbine is a small power gas turbine (a few dozen kW), usually operating
with a low compression ratio and with a regenerator to enhance performance [8].

We are interested here in

a setup of
trigeneration where the gases
coming from a micro-turbine
125 kWe are used
on one hand to provide the
heat necessary for
desorber of a machine at
absorption LiBr-H2O,
2 etc
on the other hand to produce
0.5 kg/s of hot water
sanitary at about 80 °C.
The turbine sucks 0.78 kg/s
air that is compressed to 5
bars, then goes into a
regenerator before being
brought to 950 °C in the
combustion chamber
burning natural gas. The
gas are relaxed until
the temperature of 650 °C
then cross
successively the
regenerator, the desorber
and the exchanger of
cogeneration. A
gas compressor is
necessary to enhance
the pressure of natural gas
of the GDF network. It is about
of a variant of the example
of cogeneration 8.5.1 of
[2]. Figure 14: Results of the trigeneration installation simulation

The modeling of this installation involves about thirty components, which represents several hundred.
coupled equations. All these components are available in the Thermoptim core, except for the subsystem
solution-absorbing and desorbing exchanger of the absorption refrigeration cycle, which replaces the compressor of a
In a steam compression cycle, we understand the interest presented by the mechanism of outer classes in a case like this.
that one: by adding a specialized component to represent the missing module, the work to be done is much more
reduced to if it were necessary to write a program to model the entire cycle.

It is enough to create an external component to represent the non-existent module, which on one hand engages the
couple LiBr-H2Oh, whose properties can be modeled either directly in the external component or in a body.
external individual, and on the other hand requires both a supply of high-temperature heat at the desorber level, and a
heat extraction at medium temperature at the level of the absorber. The representation of these thermal couplings is
possible by using two thermocouples, named 'Absorber' and 'Desorber' in the diagram of figure 14: the
external component "absorber-generator" calculates the thermal energies that need to be exchanged, and each
thermocouple recalculates the "exchange" transformer to which it is connected.

2.5 Gas turbine associated with a SOFC type fuel cell

 12

Several manufacturers have proposed to develop systems consisting of a gas turbine coupled with a fuel cell.
at high temperature of the type SOFC (Solid Oxide Fuel Cell), which present very high performances: for the fuel cell to
The combustor operates with high efficiency; it must be supplied with fuel and compressed air at a pressure.
around 20 bars. Only a fraction of the fuel is used, so we have, at the output of the cell, depleted air in
oxygen and reduced PCI fuel, both of which can be used in a gas turbine. Thanks to the mechanism
From external classes, it is possible to create a component that can represent the SOFC stack using an external mixer.
connected to an external divider, with the calculations being performed by it. It is thus possible to determine the variation of
chemical composition of the incoming gases and to have a model that is accurate enough to be connected to the rest of the system. The
The model of the stack associated with the gas turbine leads to the results in figure 15, for a fuel utilization rate
of 60% and a battery yield of 50%, the fuel being methane that is first steam cracked to form
hydrogen and carbon monoxide.

Figure 15: Complete diagram ( = 0.6 = 0.5)

The gases exiting the turbine being at around 550 °C, it is possible to anticipate a combined cycle that could increase the
overall yield of about 15 points, thus reaching nearly 65%. For reference, the compositions of the gases at the outlet of
The pile and before entering the combustion chamber are given in figures 16 and 17.

Figure 16: Fuel output from the pile, PCI: 5,454 kJ/kg

Figure 17: Air depleted in O at the2 outlet of the cell

2.6 Coupling to an IGCC of CO conversion units and CO capture2

As part of an Integrated Research Project of the CNRS Energy Program [9], we have coupled in Thermoptim
conversion units of CO and CO capture to a model of2 an integrated gasification combined cycle power plant
 13

(IGCC). The problem can be summarized as follows: at the gasifier outlet, we have, at a pressure of 20
bars, of a gas called 'tail gas', which typically contains 55% by volume of CO and 20% of H. CO 2 is captured
2 before the
combustion chamber, by a pressure swing adsorption (PSA) process, which allows for the treatment of a volume of
smokeless than at the power plant outlet and to take advantage of the available pressure. To do this, one must start by converting
the CO in CO. 2
 14

Figure 18: Simulation results of the two units

The complete process therefore involves two units: the first, the shift conversion unit, implements two reactors at a
different temperatures, capable of converting 85% of CO, while the second captures the CO properly 2
Say. For the shift reaction to be feasible, the tail gas must be hydrated with high-temperature steam. Finally,
 15

the captured
2 CO must be compressed into the supercritical state. The diagram in figure 18 shows the results of a simulation
carried out in Thermoptim, the two shift reactors and the PSA process capture module having been implemented in the form of
external classes.

2.7 Mise en œuvre de la méthode d'optimisation : cycle combiné à deux niveaux de

pressure

A recovery steam generator (RSG) of the type used for combined cycles is a good
example of application for the optimization method of Thermoptim. In such a GVR, steam is produced at several
pressure levels (2 to 4), whose values can be freely chosen within certain limits.

If we set aside the relatively complex theoretical support on which it is based (derived from the method of
pinching with distinction of component and systemic irreversibilities, [10], [11]), the implementation of the method
The optimization of Thermoptim is actually relatively simple to explain. It is a variant for systems.
energy methods of the Linnhoff method [12], which applies to the design of exchanger networks
complexes involving a large number of fluids as found notably in chemical engineering.

A somewhat complex energy system can involve a quite large number of fluids that exchange heat.
between them, some warming up, others cooling down. The pairing of these fluids can generally be done by
many different ways, and the choice of the best architecture is not necessarily intuitive, far from it.
To choose a high-performance heat exchanger configuration, methods derived from thermal integration emerge.
today among the most efficient, and in particular they have the advantage of providing lighting that enhances the
physical sense of the analyst while purely automatic methods force him to work blindly.

It should also be noted that the use of these methods in a tool for designing energy systems is all the more
justified that the optimization problems arising in modern high-efficiency power plants or
In cogeneration units, pinch points are specifically involved (in the recovery exchangers, at the level
of the water station...), while they did not appear in the older plants.

Figure 19: Exergetic composite curves

In practical terms, the implementation of the method can be broken down into two main phases:

The first phase consists of describing the system without making a priori assumptions about the pairings in the exchangers.
talk about unconstrained systems), and to seek to optimize the recovered energy (electric power produced, power
cogenerated...) by ensuring, through thermal integration algorithms, that there is no temperature incompatibility.
The (iterative) procedure involves varying by simulation the key parameters of the system (flow rates, temperatures, levels
depression) and to optimize its performance, while checking by the pinch method that no
additional high-temperature heat requirements while minimizing low-temperature emissions. The software provides
a modeling environment where simulation functions and the optimization method are deeply
interconnected, in order to easily vary all the characteristic parameters of the studied system.
 16

The distinction between the irreversibilities of components (specific to their own operation) and irreversibilities
systemic (related to the architecture of the system) allows us to understand the degrees of freedom that exist regarding
conception. It is during this phase that the back and forth between the people responsible takes place.
optimization and those who design the processes. One of the advantages of the method is that at any time, it is possible to
to get an idea of the challenges associated with optimization and the limits that can be reached. Classic graphical tools
Thermal integration is available, as well as the Carnot Factor Deviation Curve, which is well-suited to the problem.
posed (figure 19).

Figure 20: Combined cycle with two pressure levels

 17

- the second phase, once the system is optimized, involves searching for a compatible heat exchanger configuration (the setup
Prior implementation of the optimization method guarantees that one exists, by wisely pairing the fluids.
existing and by dividing (in series or in parallel) those that need to be. For this, Thermoptim offers blocks
exchangers that allow for step-wise operations and thus gradually define the network, by making
the pairing of fluids starting from the most constrained areas, the pinch points. It happens that constraints
technological or economic constraints require choosing a configuration of exchangers different from the one that would allow
to achieve optimal performance. It is at this stage that one realizes it.

So far, we have assumed that the interchange network was not known a priori. If it is known, Thermoptim allows
of course to model it and test it. It is thus possible to compare the initially proposed configuration with
the one that would have been achieved by an unconstrained optimization.

It is also possible to impose only part of the exchangers and optimize the rest of the network. Thermoptim then allows to
combine constrained exchangers and free exchangers, which facilitates the overall optimization of the system.

The optimization tools offered by Thermoptim are accessible from a specific screen, complementary to the one
which is used for simulation. Changes made in either of the two environments are automatically or
on order, easily transferred to the other.

This method has been implemented to solve industrial problems, at Babcock for the design of a
cogeneration installation, where it allowed for a reduction of 8% in the sizing that the Design Office had arrived at
using its classic method [13], as well as at EDF, for the optimization of cogeneration units, as well as a
282 MW gasification combined cycle power plant, where it led to an increase in electrical power of nearly
of 3.3%, or 9.4 MWe [14] (figure 19 is taken from this study).

The detailed study of a combined cycle at two pressure levels is presented in chapter 12 of [2]. It leads to
results of figure 20.

By creating a specific external component, it is possible to refine the gas turbine model presented so far.
consideration of cooling of the first expansion blades. Such work is only justified to model some
machines with high extraction or cooling by water vapor. A student from CNAM (A. Sgubini) is precisely
in the process of creating such a component to model a combined cycle with three pressure levels of the type STAG 109H
using the MS9001H turbine from General Electric.

3 Conclusion and perspectives

This note presents in a concise manner some applications of Thermoptim for the study of cycles involving
gas turbines. The software allows, as we have seen, to conduct very varied studies, from the simplest intended for
beginners to relatively complex ones, especially when it comes to integrating several techniques with each other
complementary as in the case of combined cycles and cogeneration or trigeneration units.

A version of Thermoptim, which can be described as phenomenological, has been fully operational since mid-2002.
The core components already allow for the description of a large number of energy technologies, and the mechanism
extension by addition of external classes significantly expands the potential of the tool, allowing a user to do so without difficulty.
insert specific components, at the cost of modest work and while benefiting from an environment of
secure modeling.

The current developments concern, on one hand, the sizing and non-nominal aspects, particularly for the cycles.
using gas turbines, and on the other hand control by external applications, such as training modules for
distance (FAD) or smart tutors.

We hope that by developing this type of tool, we can make access to thermodynamics easier for many engineers and
scientists who until now were deterred by the tedious nature of the calculations to be made. As soon as they can model
complex energy systems without difficulty and with great accuracy, their motivation for the discipline should
significantly strengthen.
 18

4 References
[1] GICQUEL R., Systèmes Energétiques, Tome 1 : méthodologie d’analyse, bases de thermodynamique, Thermoptim,
Presses of the School of Mines of Paris, February 2001.
[2] GICQUEL R., Systèmes Energétiques, Tome 2 : applications, Presses de l’Ecole des Mines de Paris, novembre 2001
GICQUEL R., Educational uses of Thermoptim, internal note available on request, January 2003.

GICQUEL R., Educational activities with Thermoptim: Exploration of various cycles, optimization issues,
School of Mines of Paris, LIESSE internship, October 24, 2002.

[5] TOUSSAINT J., (Coord.) Applied Didactics of Physics-Chemistry, Nathan Pedagogy, Paris, 1996.
GICQUEL R., Educational Software Tools: a response to the new challenges posed to teaching
applied thermodynamics. Example of the Thermoptim family, International TICE Conference 2000, UTC of Troyes,
18-20 octobre 2000.
CANDELIER S., Modeling of an industrial cogeneration system with Thermoptim, School of Mines
Paris, July 2001, Paris.
GICQUEL R., Trigeneration by gas micro-turbine and absorption cycle, internal note available upon request, January
2003.
TONDEUR D., Report of the PRI CO2, Energy Program of CNRS, March 2003.
GICQUEL R., Systemic optimization method based on thermal integration through the extension of the method of
pinching: application to cogeneration with steam production. General Review of Thermodynamics, Volume 34, No. 405,
October 1995.
GICQUEL R., MODICS. Generalization of the systemic optimization method to thermal systems with
imposed exchangers. General Review of Thermal Science, volume 35, pp. 423-433, 1996.

[12]LINNHOFF B., Use pinch analysis to reduce capital costs and emissions, Chemical Engineering Progress, August
1994, pp. 32 - 57
[13] LEIDE B., GICQUEL R., Systemic approach applied to dual pressure HRSG, International Gas Turbine & Aeroengine
Congress and Exhibition, Orlando, Florida, June 2-5, 1997
GICQUEL R., WILLIAMS M., AUBERT K., Optimization of the water-steam cycle of an IGCC plant, Conference
hpc'01, Paris, September 2001.

Annex 1: Presentation of Thermoptim

Thermoptim is a software package that allows for easy calculation of relatively complex thermodynamic cycles without
not obliged to resort to extreme simplifications, nor to engage in tedious calculations. It is
composed of four interconnected complementary work environments: a diagram editor (or synoptic), a
simulator, interactive thermodynamic diagrams, and an optimization method.
 19

Originally, Thermoptim was developed to contribute to

to solve certain difficulties encountered in learning and
the deepening of applied thermodynamics. Its
current features allow him to go well beyond this
the initial objective is to be used in the industry, for example for
optimize the integration of future power plants
generation involving several hundred elements.

Figure 21: Diagram of a refrigeration machine

Given that energy conversion technologies appear as assemblies of connected components

among them (compressors, turbines, combustion chambers, exchangers...), Thermoptim seeks to make the most of
synergies that exist between a relatively innovative systemic approach and an analytical and/or empirical approach
more classic:
each functional element is represented by an appropriate primitive type of Thermoptim (body, point, transformation,
node, exchanger) with modifiable specific characteristics and coupling variables;
the complete system is modeled by assembling these types through an interactive interface;
The simulation of the complete system is then managed by an automatic recalculation engine that utilizes the properties.
systemic implicitly described during modeling.

The details of Thermoptim's features can be found in [1] and [2] or from the website:
[Link]

A1.1 The four work environments of Thermoptim

The schema editor (figures 1 and 21) allows for a qualitative description of the system being studied. It includes a palette
presenting the different representable components (heat exchanges, compressors, expansion devices, chambers of
combustion, mixers, dividers...), and a work panel where these components are placed and interconnected by links
vector graphics. This graphic environment offers a particularly interesting user-friendliness for visualizing and controlling the
connections of large projects. It also allows for simplifying data entry when creating a new
project.
The simulator allows to quantify and then solve the model described qualitatively in the schematic editor. Its screen
The main gives access to the settings screens for the properties (logical and thermodynamic) of the different elements (the
primitive types of Thermoptim) involved in the generated models.

The interactive thermodynamic diagrams (figures 22 and 23) allow you to display the state with a simple mouse click.
complete thermodynamics of a fluid, thus facilitating the acquisition of the desired values. They can also be used
to visualize the thermodynamic cycles calculated by the simulator.
 20

Figure 22: Refrigeration cycle on entropy diagram Figure 23: Diagram (h, log p)

The following diagrams are available today:

1) vapor diagrams, which present the liquid zones, liquid-vapor equilibrium, and vapor of about fifteen substances
purs, where the water, with coordinate systems (depending on the body):
- the entropic diagram (T,s), with entropy on the x-axis and temperature on the y-axis;
- the (h,P) diagram called by refrigeration technicians, usually on a semi-logarithmic scale, with enthalpy on the x-axis and the
pressure on the ordinate;
- the Clapeyron diagram (P,v), for water only, with volume on the x-axis and pressure on the y-axis;
- the Mollier diagram (h,s), with enthalpy on the x-axis and entropy on the y-axis;
- the exergic diagram (h,xh), with enthalpy on the x-axis and exergy on the y-axis;
- the exergetic diagram (s,xh), with entropy on the abscissa and exergy on the ordinate.

2) diagrams of ideal gases, allowing variation in the nature of the gas (pure or compound), with systems of
coordinates:
- the entropic diagram (T,s), with entropy on the x-axis and temperature on the y-axis (open by default when the
application launch) ;
- the Clapeyron diagram (P,v), with volume on the x-axis and pressure on the y-axis.

3) diagrams of wet mixtures (mixtures of water and an ideal gas), allowing to vary on one hand the nature
dry gas (air, but also pure gases or compounds, including combustion gases), and on the other hand the pressure, with as
coordinate systems:
- the psychrometric diagram or Carrier diagram (w,t), with the dry bulb temperature on the x-axis and specific humidity
in ordinate;
- the Mollier diagram wet (w,q') or (w,h), with specific humidity on the x-axis and specific enthalpy in
ordered.

The fourth working environment of Thermoptim is its optimization method through systemic integration, which is
an extension in the case of energy systems (electricity generation plants, cogeneration...) of the method of
Pinching developed within the framework of chemical engineering to optimize the configuration of very large networks of exchangers.
(such as those of a refinery).

In Thermoptim, the couplings between these four environments are achieved through particular interfaces, integration
the most advanced being carried out between the simulator and the schematic editor on one hand, and between the simulator and the method
of optimization on the other hand.

The main strengths of Thermoptim are as follows:

it provides a consistent modeling environment combining the contributions of systemic and analytical approaches.
With a relatively limited initial investment in time to understand the underlying logic, it itself
quite natural and intuitive, it allows for a very rigorous and efficient modeling approach. The
developed models can be easily verified, documented, archived, and reused or modified;
its base of primitive types constitutes a grid for analyzing energy technologies, allowing to highlight
the graph of the interrelations of the components. With a bit of experience, the modeler can easily
 21
represent an energy technology in a way that is both faithful to reality and computable by the software, knowing that
if necessary, he can add his own components to those of the core;
the description of the problem is made using the natural concepts of the engineer or the physicist, the types
primitives with a very clear physical meaning. The modeler can thus focus on the physical analysis of
phenomena, the mathematical, numerical and IT translation of the generated model being ensured by the tool of
transparent manner;
If he uses the core components of the software package, the modeler does not have to write a line of code or solve it himself.
even an equation, Thermoptim ensuring the computer translation of the project description submitted to it.
The user is thus assured of obtaining realistic results quickly without risks of programming errors. If he
wants to perform calculations not provided for in the tool, he has the leisure to do so by exploiting (for example with macros in
a spreadsheet) the text file of the results of his project, or by adding his own post-processing modules
external
Given that the properties of fluids are calculated automatically, the modeler can obtain results.
very precise while being relieved of many computational difficulties that he could not have avoided otherwise
resorting to caricatured simplifying assumptions.

The Thermoptim simulator calculates the different elements of a project step by step. It is a sequential calculation mode.
which distinguishes it from other modeling environments (matrix-based) in which the entire set of problem equations
is resolved simultaneously. This way of doing has the advantage that it is much easier to calculate
successively the elements one by one than to solve the complete system all at once. On the other hand, it leads to two
difficulties: on one hand, it may be necessary to iterate the calculations several times to find the correct solution,
notably if the system is coupled, and on the other hand, for a somewhat complex project, the question arises of knowing in which
the order in which the calculations must be performed. To solve this latest difficulty, a set of algorithms has been developed.
Called the automatic recalculation engine of Thermoptim, it is a key component of the Java version of the software package.
The specific screen allows you to track the steps of recalculations, ensuring the relevance of the modeling.

A1.2 Core extension mechanism by adding external classes

The interfacing of Thermoptim with external classes (Java code elements) facilitates interoperability.
software with the outside world, especially with other applications developed in Java. The interest is twofold:
to be able to create Thermoptim extensions from the common core by adding external modules recognized by
the software package, which defines elements that appear on its screens transparently for the user. It is
as well as a user can add their own bodies or components not available in the core;
to be able to control Thermoptim from another application, either to guide a user (intelligent tutor, module of
remote training), either to monitor the execution of the code (management or regulation, access to libraries
thermodynamics)
 22

A1.3 Exploration of models

Figure 24: Catalogue of examples

Afin de faciliter les activités d'exploration de modèles qui constituent l'une des principales catégories d'utilisation
Thermoptim's educational tool includes an automatic procedure to search for and load modeled examples with
the software package. You access the window shown in figure 24 by selecting the line "Examples Catalog" from the "Files" menu.
from the simulator screen or by typing Control E. It shows the list of different available examples, in this figure
grouped by chapters of the referenced book in [1] and [2]. By selecting one of them, a brief appears
presentation of the example, which allows to ensure the choice well. A double-click on the selected line or a click
Clicking 'Open' triggers the loading of the project file as well as the schema file if it exists. In the case of the figure
24, this is the example of the simple steam cycle referenced under number 6.1.1 in the chapter 'Production Cycles'
steam electricity.

Each teacher can easily create their own catalog of examples if they wish.

A1.4 Model Building

Building a thermodynamic system model with Thermoptim is very simple and is done in two steps:
we start by making a qualitative description of it by representing it graphically as a set of
components (more generally functionalities) connected together by vector links representing the
fluid pipelines or heat exchangers.
the model thus defined is then quantified by parameterizing the different primitive types it involves, in order to be able to
calculate the performance

The schema editor allows for the qualitative step: the user initially provides only the minimum.
necessary information for the logical definition of the project it builds (implicitly the types of components it
select, and explicitly their name and that of the exit point and the body associated with them, as well as the value of
flow that flows through them). Then, when he connects these components together, some of this information is
automatically propagated from upstream to downstream (for example, the entry point of the downstream component becomes the same as
the output of the upstream component). This small amount of basic information provided to the graphic modeler corresponds to
the systemic description of the studied project.

Once this step is completed, it becomes possible to transfer the components of the diagram into the simulator, to create the
required primitive types, with a default configuration of their thermodynamic properties. The quantification of the model
Desire can then be made by refining this setting, each element of the simulator displayed very simply in double-
clicking either on the corresponding component in the schematic editor or on the line of a table in the main screen.
 23

Once the elements are set up and their calculation is complete, the obtained results can be directly displayed in
the schema editor that thus becomes a true overview of the installation, or represented in the form of a cycle
in one of the interactive diagrams.

A1.5 Structure of Thermoptim P R O JE T

P o ints T r a n s f o r m a t I or n s
C o rp s p o in t a m o n t p o in t a v a l

Noeuds E c h a n g e u rs

tra n s fo s d e s b ra n c h e s tra n s fo c h a u d e t r a n s f o f r o Id e
Figure 25: Structure of Thermoptim

We have seen that the structure of Thermoptim is based on the distinction of a certain number of primitive types:
The concepts of bodies and points allow for the calculation of the state and intensive properties of the fluids used, for various
valeurs de la pression, de la température.
these fluids undergo evolutions (transformations or transforms) that can be grouped into a few major
categories, the most common of which are the following: compressions, relaxations, combustions, and exchanges
of heat. It is at this level that the mass flows involved are specified, and therefore the extensive properties
may be calculated.
the fluids involved flow through the machines forming more or less complex networks that need to be managed.
describe. The transformations highlighted previously correspond to part of these circuits. For the
to complete, it is necessary to call on nodes (dividers, mixers, or separators).
When two fluids mutually exchange heat, they form heat exchangers, coupled components.
where the two transformations cannot be calculated separately.

All these elements (bodies, points, transformers, nodes, exchangers) are grouped in the software package under the name of
project. Figure 25 shows the logical relationships that exist between these primitive types. It helps students to structure the
representation they have of these different notions.
 24

1 General introduction1
1.1 Educational applications1
1.2 Industrial applications2
1.2.1 Facilitate and secure the modeling of energy systems3
1.2.2 Optimize complex energy systems using exergy methods.3
2 Illustrations of the use of Thermoptim for the study of gas turbine cycles.4
2.1 Simple gas turbine model (learning)..........................................................................................................4
2.1.1 Development of the model4
2.1.2 Energy and Exergetic Balance5
2.1.3 Related Educational Activities6
2.2 Combined cycle at a pressure level7
2.3 Installation of industrial cogeneration8
2.4 Installation of micro-turbine trigeneration9
2.5 Gas turbine associated with a SOFC type fuel cell.10
2.6 Coupling to an IGCC of CO conversion units and CO capture11 2
2.7 Implementation of the optimization method: combined cycle with two pressure levels.13
3 Conclusion and perspectives15
4 References15
Annex 1: Presentation of Thermoptim16
A1.1 The four working environments of Thermoptim17
A1.2 Mechanism of core extension by adding external classes....................................................................18
A1.3 Exploration of models19
A1.4 Model Construction19
A1.5 Structure of Thermoptim20