Puigjaner L. (ed.) Syngas from Waste - Emerging Technologies
Подождите немного. Документ загружается.
11. Diwekar U (2005) Green process design, industrial ecology, and sustainability: a systems
perspective. Resour Conserv Recy 44:215–235
12. Diwekar U, Small M (2002) Process analysis approach to industrial ecology. In: Ayres RU,
Ayres LW (eds) A handbook of industrial ecology. Edward Elgar, Cheltenham, UK, pp 114–
137
13. Nocedal J, Wright S (2006) Numerical optimization, 2nd edn. Springer, Berlin
14. Steuer R (1986) Multiple criteria optimization: theory computation and application. Wiley
series in probability and mathematical statistics–applied, 1st edn. Wiley, New York
15. Statnikov R, Matusov J (1995) Multicriteria optimization and engineering, 1st edn. Chapman
and Hall, New York
16. Griva I, Nash G, Sofer A (2009) Linear and nonlinear optimization, 2nd edn. Society for
Industrial and Applied Mathematics (SIAM), Philadelphia
17. Caballero J, Odjo A, Grossmann I (2007) Flowsheet optimization with complex cost and size
functions using process simulators. AIChE J 53:2351–2366
18. Biegler L, Cuthrell J (1985) Improved infeasible path optimization for sequential modular
simulators, Part II: the optimization algorithm. Comput Chem Eng 9:257–267
19. Lang Y-D, Biegler L (1987) A unified algorithm for flowsheet optimization. Comput Chem
Eng 11:143–158
20. AspenTech INC (2005) Aspen plus system reference. Aspen Tech, Cambrige
21. AspenTech INC (2005) Aspen HYSYS system reference. Aspen Tech, Cambrige
22. Alexander B, Barton G, Petrie J, Romagnoli J (2000) Process synthesis and optimisation tools
for environmental design: methodology and structure. Comput Chem Eng 24:1195–1200
23. Chen H, Rogers T, Barna B, Shonnard D (2003) Automating hierarchical environmentally-
conscious design using integrated software: VOC recovery case study. Environ Prog 22:147–
160
24. Caballero J, Milan-Yanez D, Grossmann I (2005) Rigorous design of distillation columns.
Integration of disjunctive programming and process simulators. Ind Eng Chem Res
44(200):6760–6775
25. Diwekar U, Grossmann I, Rubin E (1992) An MINLP process synthesizer for a sequential
modular simulator. Ind Eng Chem Res 31:313–322
26. Chaudhuri P, Diwekar U (1997) Synthesis under uncertainty with simulators. Comput Chem
Eng 21:733–738
27. Fu Y, Diwekar U, Young D, Cabezas H (2001) Designing processes for environmental
problems. In: Sikdar SK, El-Halwagi M (eds) Process design tools for the environment, vol 1,
Chap. 12. Taylor & Francis, New York, pp 295–317
28. Fang K, Li R, Sudjianto A (2006) Design and modeling for computer experiments, 1st edn.
Chapman & Hall/CRC, Taylor & Francis, Boca Raton, FL
29. Chen J, Frey C (2004) Optimization under variability and uncertainty: a case study for
NOx emissions control for a gasification system. Environ Sci Technol 38:6741–6747
30. Almeida-Bezerra M, Santelli R, Padua-Oliveira E, Escaleira L (2008) Response surface
methodology (RSM) as a tool for optimization in analytical chemistry. Talanta 76:965–977
31. Kasabov N (1998) Foundations of neural networks, fuzzy systems, and knowledge
engineering, 2nd edn, a Bradford book. MIT Press, Cambridge, Massachusetts
32. Pérez-Fortes M, Bojarski AD, Velo E, Nougués JM, Puigjaner L (2009) Conceptual model
and evaluation of generated power and emissions in an IGCC plant. Energy 34:1721–1732
33. Pérez-Fortes M, Bojarski AD, Velo E, Puigjaner L (2010) IGCC power plants: Conceptual
design and techno-economic optimization. In: Harris AM (ed) Clean energy: resources,
production and developments. NOVA, Hauppauge, NY
34. Giampaolo T (2006) The gas turbine handbook, principles and practice, 3rd edn. Fairmont
Press, Lilburn, Georgia
35. Zhu Y (2004) Evaluation of gas turbine and gasifier-based power generation system. PhD
thesis, Graduate Faculty, North Carolina State University
36. Ongiro AO, Ugursal VI, Altaweel AM, Blamire DK (1995) Simulation of combined-cycle
power-plants using the Aspen Plus shell. Heat Recovery Syst CHP 15:105–113
198 A. D. Bojarski et al.
37. Coca MT (2003) Tecnología de gasificación integrada en ciclo combinado: GICC. Aplicación
real en España: Elcogas. Puertollano. Technical report, ENERCLUB
38. Highman C, van der Burght M (2003) Gasification. Elsevier Science, Amsterdam
39. Wells G, Rose L (1986) The art of chemical process design, Vol 1–2 of computer-aided
chemical engineering. Elsevier, Amsterdam
Modelling Superstructure for Conceptual Design 199
Process Integration: HEN Synthesis,
Exergy Opportunities
Zdravko Kravanja, Miloš Bogataj and Aleksandr Soršak
Abstract This chapter provides a brief description of thermodynamic analysis,
the maximization of heat-recovery and power generation and the synthesis of
IGCC’s heat exchanger network. Trade-offs between the income from power
generation and utility costs plus the investment in HEN are studied with respect to
the generation of different pressure-levels of steam and temperature driving force
losses within the heat-recovery network. A combined pinch analysis/mathematical
programming approach is applied and the optimization models are described for (i)
the maximization of heat-recovery and power generation, and (ii) a synthesis of
the heat-recovery and steam/power generation network. A sensitivity analysis
for the synthesis is performed in order to show how optima are sensible for the
expected increase in future electricity and utility prices, and the project’s lifetime.
The results obtained from the studied IGCC process indicate that detailed opti-
mization has to be performed during the network synthesis step, otherwise optimal
trade-offs are missed that may result either in serious power generation losses or in
obtained over-designed networks.
Z. Kravanja (&) M. Bogataj
Faculty of Chemistry and Chemical Engineering, University of Maribor,
Smetanova ulica 17, 2000 Maribor, Slovenia
e-mail: kravanja@uni-mb.si
M. Bogataj
e-mail: milos.bogataj@uni-mb.si
A. Soršak
Javni zdravstveni zavod Mariborske lekarne Maribor,
Minarikova ulica 6, 2000 Maribor, Slovenia
e-mail: sandi.sorsak@mb-lekarne.si
L. Puigjaner (ed.), Syngas from Waste, Green Energy and Technology,
DOI: 10.1007/978-0-85729-540-8_9, Springer-Verlag London Limited 2011
201
Notation
ASU Air separator unit
CP Claus plant
GCC Grand composite curve
GT Gas turbine
HE Heat exchanger
HEN Heat exchange network
HPS High-pressure steam
HRAT Heat-recovery approach temperature
HRSG Heat-recovery steam generator
LP Linear programming
LPS Low-pressure steam
MINLP Mixed-integer non-linear programming
MPS Medium pressure steam
OA/ER Outer-approximation/equality-relaxation
ST Steam turbine
1 Introduction
In this chapter we demonstrate on the integrated gasification combine cycle (IGCC)
case-study how to perform energy efficiency enhancement analysis and synthesis of
heat/energy-recovery networks. Either heuristic, pinch analysis or mathematical
programming can be applied to fulfil the task. In our case, a combined pinch analysis/
mathematical programming approach is described. The enhancement analysis and
synthesis are based on the assumption of considering process streams with fixed
flows, temperatures, pressures, phase conditions and composition. If so, then, at this
stage, we skip the gas turbine (GT) system by focusing only on the enhancement of
IGCC’s process heat-recovery network for steam/power generation.
Basic features of the IGCC process in regard to heat integration and the
production of steam for power generation are the following:
• We are dealing with a large number of process hot and cold streams (about 40)
within a wide range of temperatures (100–1,270 K).
• Low-pressure (LP-), medium-pressure (MP-) and high-pressure steam (HPS)
can be produced for power generation, presented as additional hot and cold
streams (16) with variable heat capacity flowrates.
• The heat to be transferred between hot and cold streams is extensive (about
400 MW), giving rise to very large areas of heat exchangers.
• Approximately one third of all the process and utility streams are isothermal.
• On the one hand we have a large heat-recovery steam generator (HRSG) for the
production of various steams, and on the other hand, we have the generation
of power within a three-stage steam turbine system (vapour turbine—VT)
operating at a high, medium and low pressure.
202 Z. Kravanja et al.
It should be noted that the exergy content of steam, and, hence its potential to
produce power, increases with pressure level and temperature. Thus, with HPS we
can produce more power than with MPS, and with MPS more than with LPS. It is
then apparent that, for the maximization of power generation, both the amount of
heat recovered from process hot streams (the load) and steam levels are equally
important. However, there is a general trade-off between the load and the level since
the higher the level, the less heat can be recovered, and vice versa. Another very
important trade-off is the one between the investment in HEN and the amount and
level of heat recovered. A smaller heat-recovery approach temperature (HRAT)
enables better integration; however, it is achieved at a higher investment cost. On the
other hand, a larger HRAT would recover less heat into steam at a lower exergy
level, at the same time wasting more heat and using more cold utility. Another
important question relates to the selection of appropriate types of heat exchanger
(HE) units. In order to exploit the temperature potentials of process hot streams as
much as possible, in such a load versus level problem, it is very important to consider
those HE types that exhibit counter-flow or a very near counter-flow arrangement of
hot and cold streams. The selection of different hot and cold utilities is also an
important issue. The higher the level of a hot utility and the lower the level of a cold
utility below the ambient temperature, the higher are the costs. And lastly, due to the
economy-of-scale effect, investment costs for equipment are generally lower with a
smaller number of process units, so that it is also necessary to design a HEN with a
smaller number of HE units. It should also be noted that one has to consider a parallel
arrangement of HE units when a calculated area exceeds the maximal allowable
areas for selected types of HE units.
With respect to the above-mentioned trade-offs, it is apparent that we are
dealing with a very complex optimization problem. Therefore, the main goal is to
obtain a solution where optimal trade-offs are established between the income
from power generation and the outcome from the consumption of utilities plus
investment. Such an optimal HEN design should exhibit an optimal generation of
different quality steam (load and levels) at optimal temperature driving forces in
the optimal arrangement, number and types of HE units. In order to accomplish
this task, the following step-wise procedure is applied:
1. Thermodynamic analysis of the IGCC process flowsheet in order to assess the
opportunity for heat and exergy integration for heat-recovery and power
generation.
2. Maximization and sensitivity analysis of heat-recovery and power generation
versus HRAT.
3. Optimal synthesis of a heat-recovery and steam/power generation network.
Whilst in the first step a thermodynamic approach with pinch analysis is
applied, the second and third steps rely on mathematical programming: the second
step on linear programming (LP) applies an extended model of the simultaneous
heat integration model by Duran and Grossmann [1], and the third HEN synthesis
step on mixed-integer non-linear programming (MINLP) is based on an extension
of the simultaneous model for the synthesis of HEN by Yee and Grossmann [2].
Process Integration: HEN Synthesis, Exergy Opportunities 203
2 Thermodynamic Analysis of the IGCC Process
A very important step when performing thermodynamic analysis of any process is
its data collection and acquisition. Let us first describe how data is extracted from
the IGCC flowsheet.
2.1 Data Extraction
The extraction of process streams’ thermodynamic properties plays an important
role in the heat integration analysis of a given process. Regardless of whether the
analysis is performed using the pinch methodology or any of the approaches based
on mathematical programming, great care must be taken to extract the data
properly. By this we mean that the data should be extracted in a way that imposes
no unnecessary constraints on the problem being analyzed and yet ensures feasi-
bility of the final design.
Let us now consider the schematic representation of an IGCC process in Fig. 1.
The figure is a condensed representation of a converged Aspen Plus flowsheet [3].
In each of the blocks of the flowsheet (GT, ASU, G, CP, ST and GCT), the
represented streams were identified as those with the need for cooling or heating in
order to satisfy process constraints. Stream names correspond to the names of the
process units in the flowsheet.
As shown in Fig. 1, some of the streams undergo a phase change. These streams
are denoted with shaded rectangles, for example, condensation of the top product
in the distillation column (REGEN-COND in GCT). On the other hand, some
Fig. 1 Schematic
representation of enthalpy
streams in IGCC process
204 Z. Kravanja et al.
streams actually represent a single stream, although extracted as multiple streams
with linked target and supply temperatures. Such a stream may even cross the
borders of blocks. These streams are denoted with dotted lines. For example,
stream N2HEATER in GT is linked to stream AC12 in the same block, which is
then linked to stream ABSORP in ASU.
The above two cases are the most common examples which, unless the stream
temperature-enthalpy profiles are represented correctly, can lead to errors and
missed opportunities. Firstly, let us consider the case when a phase transition takes
place on a generic cold stream, as depicted with a solid line in Fig. 2. The stream
with a supply temperature of 295 K is being heated up to boiling temperature
(373 K), evaporated, and then superheated to a target temperature of 400 K. The
total enthalpy flowrate equals 1 MW. Please note that since we are dealing with a
continuous process, enthalpy flowrates (I/W) are used rather than enthalpies (H/J).
Obviously, we are dealing with three segments, each of them having a unique heat
capacity—slope of the temperature-enthalpy profile. Now, suppose that the stream
is extracted as a single segment stream i.e., having constant heat capacity through
the entire temperature-enthalpy profile (dashed line in Fig. 2).
Clearly, as it can be seen from this figure, the latter case causes thermodynamic
infeasibilities due to negative temperature differences in a portion of the profile
where the hot stream (dotted line) is colder than the properly represented cold
stream. For this reason, whenever a phase transition occurs in a given stream, its
temperature—enthalpy profile must be represented by segments. The number of
segments of each stream, their target and supply temperatures and enthalpy
flowrates can be obtained by simulating each such stream, as shown in Fig. 3.
Next, we must address those streams extracted as multiple streams, but are
actually a single physical stream. Unlike the prior example, where improper
representation (extraction of thermodynamic properties) of streams can lead to
serious errors, this is not the case here. Nonetheless, extracting the data in this way
does not open up additional opportunities for improving the design. The latter
comes from the fact that each such stream imposes additional constraints (fixed
supply and target temperatures, and enthalpy flowrate) on the problem being
Fig. 2 Temperature-
enthalpy profiles of a stream
subjected to phase transition
Process Integration: HEN Synthesis, Exergy Opportunities 205
analyzed, thus reducing the degrees of freedom. Obviously, it is a must that such
streams are merged into a single stream having a supply temperature equal to the
supply temperature of the first stream in a series, a target temperature equal to the
target temperature of the last stream in a series and an enthalpy flowrate equal to
the sum of the enthalpy flowrates for all the streams in a series.
Having discussed the above issues, let us take a closer look at the raw data in
Tables 1 and 2. In Table 1, the basic thermodynamic properties of hot streams are
given as extracted from the flowsheet. The properties of cold streams are given in
Table 2. In total, 22 hot and 15 cold streams were identified. Among the hot and
cold steams, the streams found in the GT block (pre-heating of condensate in B14,
Fig. 3 Flowsheet for
extracting the thermodynamic
properties of streams
undergoing phase transition
Table 1 Thermodynamic
properties of hot streams (raw
data)
Block Process
unit/stream
T
in
/(K) T
out
/(K) I/(MW)
1 ABSORP 400.1 287.1 16.397
1 HECR1 287.1 110.6 33.578
1 HECR2 370.1 123.3 16.268
2 QHP 1,108.1 658.1 72.210
2 QIP 658.1 508.1 22.797
3 COND2 383.1 382.1 0.559
3 COOLER 384.7 313.1 0.536
3 HCLEAN1 415.4 382.0 6.721
3 HCOOL 382.0 306.1 12.865
3 HE2 389.1 344.0 5.668
3 COOLER 344.0 306.1 4.720
3 REGEN
COND
357.5 356.5 16.785
4 BOILER 1,273.1 543.1 7.602
4 B4 573.1 349.2 1.529
4 B6 777.6 414.1 4.728
5 HXHP 849.3 599.3 146.475
5 HXIP 599.3 449.3 83.740
5 HXLP 449.3 443.1 3.392
5 HXCOND 443.1 373.1 38.143
6 N2HEATER 623.9 563.1 9.050
6 AC12 563.1 400.1 23.818
7 B7 307.2 306.2 276.700
Blocks: 1 ASU, 2 G, 3 GCT, 4 CP, 5 HRSG, 6 GT, 7 ST
206 Z. Kravanja et al.
production of steam at different pressures in HXHP, HXIP and HXLP and con-
densation of turbine exhaust in B7) are also given, although the steam is generally
treated as a utility and should, therefore, be left out of the analysis. However, in
this case, it may be advantageous to treat these streams as hot and cold streams,
since they directly contribute to the production of the final product (electrical
power).
The raw data given in Tables 1 and 2 carries all the above-discussed pitfalls.
Therefore, the data should be refined further in accordance with the above dis-
cussion. All of the streams which undergo a phase transition should be checked as
to whether they should be broken up into multiple segments. In addition, some
streams should be merged and represented as a single stream. The refined data is
given in Tables 3 and 4. Besides the information on the supply and target tem-
peratures and enthalpy flowrates, additional data needed for the HEN synthesis
approach, which is described and discussed in subsequent sections, is also pre-
sented in these tables. As a result of the merging and breaking-up of streams, we
end-up with 19 hot streams (16 process and 3 steam-condensing streams) and 22
cold streams (13 process and 9 steam-producing streams). Note that for the sake of
further analysis steam-condensing and steam-producing streams will be treated
with variable heat capacity flowrates. Therefore, specific capacity flowrate, defined
in MW per mass production rate of steam in ton h
-1
, is introduced for those
streams.
Last, but not least, it must be pointed out here that the streams given in
Tables 3 and 4 are not all the process streams, found in the flowsheet, which
could also be considered in the analysis. For practical reasons, half of dozen or
so streams having enthalpy flowrates well below 0.5 MW were neglected in
this study.
Table 2 Thermodynamic
properties of cold streams
(raw data)
Block Process
unit/stream
T
in
/(K) T
out
/(K) I/(MW)
1 B20 110.6 510.6 33.502
1 B15 121.6 661.4 15.985
1 B1 110.6 111.6 10.403
3 REBOILER 408.1 409.1 0.597
3 TOHYDRO 397.6 414.1 1.271
3 REBOILER2 408.1 409.1 0.814
3 HE1 328.9 371.1 5.668
3 REGEN
BOIL
389.0 390.0 25.195
3 HCLEAN2 342.4 406.8 6.938
6 N2HEAT2 846.6 966.0 9.050
6 CLGHEATER 397.7 533.1 10.778
7 HXHP 413.1 764.1 218.959
7 HXIP 413.1 567.8 106.579
7 HXLP 413.1 518.0 3.392
7 B14 306.2 413.1 56.413
Blocks: 1 ASU, 2 G, 3 GCT, 4 CP, 5 HRSG, 6 GT, 7 ST
Process Integration: HEN Synthesis, Exergy Opportunities 207