close

Вход

Забыли?

вход по аккаунту

?

Use of Simulation and Expert Systems to Increase the Energy Efficiency in Cane Sugar Factories.

код для вставкиСкачать
Dev.Chern Eng Mineral Process., lo(In),pp. 145-179,2002
Use of Simulation and Expert Systems to
Increase the Energy Efficiency in Cane
Sugar Factories
Osvaldo Goza Leon*, HCctor PCrez de Alejo Victoria and
Marcel Rijckaert'
Sugar Research Group, Faculty of Chemical Engineering, ISPJAE,
Marianao, Havana City, Cuba.
'Expert Systems Applications Development
Engineering Department, KU Leuven, Belgium.
Group,
Chemical
As part of a strategy to increase the energy eficienq of Cuban sugar factories, an
automated method of analysis has been developed and is being applied based on the
use of STA and ANSTE programs STA is a steady state sequential moduiar
simulation program aimed at evaluating energv systems in cane sugar factories
ANSTE which is an expert system based on calculations with STA and certain
additional information, makes it easy to detect problems affecting the energy
efJicienq of the factory, and to make recommendations for solutions In this paper,
the strategy of these two programs is briefly discussed, and the main features of the
STA and ANSTE programs are presented Some problems detected by the expert
system and the conceptions of its knowledge base are described An application in
an actual factory is presented In this factory, both programs were applied in the
evaluation of the energy system, and proposed solutions were accepted by managers
Introduction
Process analysis is a very effective tool to increase the technical and economic
efficiency of chemical industries, with the increasing use of computers techniques.
Considering the existence of 156 sugar factories in Cuba, and the complexity of their
thermal schemes, process simulation becomes a very valuable tool to evaluate the
performance of these factories.
Simulation programs provide a large amount of information because they generate
multiple alternatives and take many factors into account This is why it is desirable to
search for automated ways to analyze all of this information. Energy systems in cane
sugar factories have several features that make their analysis particularly complex:
* Author for correspondence (e-mail
ogoza@yahoo com)
165
0 God Ledn, H Pe‘rez de Alejo and M Rtjckaerl
Behavior of the factory is strongly dependent on the characteristics of the feed
raw material, the cane, which may vary over a wide range
0
There is a close link with the technological process, which demands
considerable amounts of heat and power.
0
These systems are highly interactive, with many different pieces of equipment
and configurations.
On the other hand, the cane sugar industry has some features that make it very
attractive in the search of greater efficiencies. The feed raw material brings the fuel
needed, the bagasse, which will not only be able to satisfy the energy needs of the
factory but also have a surplus to be used for other purposes.
Cane sugar factories offer unique opportunities for the application of cogeneration
technology. Depending upon cane quality, factory configuration and the degree of
steam economy practiced, surplus bagasse can be produced for conversion into
secondary products, commercial power generation being usually in great demand and
normally very profitable. The immediate result of producing power fiom sugar cane is
an improvement in revenue for the producer - a legitimate reward. But the long-term
result will benefit the entire community, every kilowatt-hour of electricity produced
from bagasse and exported to the grid reduces the use of fossil fuel, which in turn will
reduce atmospheric pollution and global warming, and replace the world‘s rapidly
declining fossil fuel reserves with a renewable energy resource. The main reason why
cane sugar factories have such great potential for producing electricity is that the
sugar manufacturing process is ideally suited for the sequential use of ,steam of
different pressures, a highly efficient practice which greatly reduces the energy
requirements for sugar production and releases steam for production of electricity.
Under ideal conditions, a sugar factory can export as much as 80% of the energy it
can generate ltom bagasse [ 11.
Historically, broad indices have been used to provide a general picture of the
energy efficiency of a sugar factory cogeneration system. But more precise
measurement of the sugar factory energy efficiency can boost financial returns when
sugar factories start to sell excess electricity. However, as stated in [2], there are no
recognized methodologies for measuring and comparing this efficiency amongst sugar
mills. This has been the main reason why the International Cane Energy Network
commissioned the preparation of the Sugar Factory Energy Data Protocol [3] which is
still being validated. Central to this effort is the identification of key indices for use in
reporting raw-sugar factory energy consumption and energy efficiency in a
standardized manner.
“The need to make better use of energy in sugar factories and the factors
mentioned above have been the reasons to elaborate on, and to present a general
strategy of analysis for energy systems in cane sugar factories, and to implement an
automated method of analysis which combines the use of process simulation and an
expert system” [4].
0
Automated Method of Analysis
The method of analysis of energy systems in cane sugar factories has been conceived
in relation to a general strategy that has the following steps:
Characterization of the factory
166
Increasing the Energy Eficiency in Cane Sugar Factories
Identification of problems that affect the energy efficiency.
Statement of alternative solutions.
0
Technical and economical selection of best alternatives,
0
Implementation and checking of the chosen alternatives.
This strategy requires vast and diverse information. It should consider multiple
alternatives which would allow it to establish behavior trends of the system, and in
conjunction with economic analysis to establish the best ways to increase the
efficiency of the energy system.
Once the factory has been characterized, the method used to apply the strategy to a
specific operating case (one running), consists of the steps shown in Figure 1.
0
Mass and energy
balances
1
Calculation of
indicators
3
Process Simulation
CSTA)
Expert System
cJ===)
Recommendation
of solutions
Figure 1. Automated method of analysis
Both the STA simulation program (1979) and the ANSTE expert system (1996)
have been developed by our Sugar Research Group (Faculty of Chemical
Engineering, Havana) headed by the authors, and they have been continuously
improved. This paper aims at presenting the main features of both programs and its
combined application, which have proven very useful for the Cuban sugar industry.
STA Simulation Program
Process simulation is a powerful tool that is being extensively used by advanced
chemical industries to maintain and increase their competitiveness [ 5 ] , even more so
when combined with other types of software such as expert systems and process
integration programs [6].When solving energy-saving related problems, it has been
recognized that the applications of simulation are as broad as the fields of energy,
167
0 Coai Leon, H Pe‘rez de Alejo and M.Rijckaert
being especially useful for the formulation of energetic strategies that take into
account the varying costs of the energy and the varying costs of products.
The development of commercial simulators has not been focused on the sugar
industry but on other industries like petrochemical and pulp and paper industries. In
this regard, we can say that when ASPEN PLUS was tested for application to CSR
refmeries, few models were available and much effort was needed to develop
customized models 171. Another relevant simulator like HYSYS lacks models
oriented to the sugar industry [8]. To our knowledge, none of the existing commercial
simulators use an expert system for the analysis of all the calculation results obtained
in the simulation of an entire factory.
Some simulation programs have been developed for the sugar industry. Among
them are the one described in [9] for energy optimization in beet sugar factories,
CANEPRO [http://www.winrock.org/reep/canepro.htm]for simulation of financial
and economic performance of sugar mill cogeneration schemes, and Sugars [lo]
which is the worlds most widely used sugar process simulator. Since none of these
simulation programs have been available in Cuba, Cuban specialists have developed
some applications for the sugar industry, which is our main industry; some of the
best-known ones are SIMFAD and STA [ 111.
The origin of the STA simulation program dates back to late 1970s [ 121, and it has
been continuously improved ever since STA (Sistema Termo Amcar) is a steadystate sequential modular simulation program, the only one of its kind developed in
Cuba. It is aimed at evaluating energy systems in cane sugar factories, both raw sugar
and refined sugar [13]. Owing to the sequential modular approach, it has a great
flexibility to represent different schemes in the information flow diagram, and the
possibility of obtaining a great amount of information both of the streams and of the
equipment [141.
The simulation is based on the information flow diagram, which is built 60m the
process flow diagram and constitutes the way the communication between user and
program is established. The input information, provided by a data file, consists of the
calculation order for modules, the parameters for input and recycle streams, and the
parameters for calculation modules. The parameters of an information stream consist
of 10 variables whose meaning are shown in Table 1.
STA has an executive program and a group of specific subroutines related to the
calculation modules, and the evaluation of physical and thermodynamic properties
(enthalpy, density, heat capacity, boiling point elevation) of sugar streams (juice,
syrup, liquor, molasses, crystalline sugar) and water (steam tables).
Available calculation modules allow STA to simulate heaters, evaporators, pan
station, steam turbines (back pressure, condensing, extraction condensing), steam
machines, mills (power), desuperheaters, steam generators, de-aerators, flash
evaporation, mixers and splitters, melters, dissoluters, driers, centrifbgals and pans for
refined sugar. One piece of equipment, for instance, a steam generator, may be
simulated by two different modules since different calculation methods based on
different concepts are used. The user may choose the most suitable module so as to
obtain the information desired. There are “auxiliary” modules aimed at obtaining
suitable reports (typical for the factory, energy and exergy related) and at forcing
convergence (direct substitution, regula-falsi methods).
168
Increasing the Energy Eficiency in Cane Sugar Factories
Table 1. Listing ofparameters of an information stream
Number
1
2
3
4
5
6
7
8
9
10
Parameter
Number of the stream in the information flow diagram
Identificationflag
Total massflow
Temperature
Pressure
Waterjlow
SucroseJlow
Non-sucrose soluble solidsflow
Fiber jlow
Insoluble non-fiber solidsflow
The executive program, with a sequential approach, performs calling of the
modules and the handling of stream information and modules parameter information
Calculations carried out by the modules include the parameters of its output
information streams, and in the case of modules representing equipment, certain
indicators such as those shown in Table 2.
Table 2. Indicatorsfor equipment
Equipment
Heaters
Evaporators
Steam turbines
Steam boilers
Parameters
Juice velocity in tubes
Heat transfer coefficient
Thermodynamic effectiveness
Inlet and outlet temperature differences
Evaporation rate
Economy
Heat transfer coefficient
Temperature difference between heating steam and juice
Thermodynamic efficiency of expansion
Mechanical and electrical efficiencies
Ideal and real specific steam rates
Thermal efficiency, Steam generated per fuel consumed
The results of the calculations, given by the complete solution of mass and energy
balances and the evaluation of indicators for equipment, may be obtained in different
presentations for both display and printer, and saved in data files.
The sequential modular approach enables the STA program to be open for future
improvement, and to have a great flexibility to represent different schemes both in
raw sugar and refined sugar factories. In this way, process integration studies of these
industries are performed quickly and rigorously
169
0 Gozd Lebn, H Pe‘rez de Alejo and M. Rijckaert
There have been versions of the STA program in Fortran, QuickBasic, Turbo
Pascal and Borland Pascal. At present, a version in Borland Delphi is being
developed, including all of the benefits to a user that are available in the Windows
environment such as a full graphical interface and database storage.
The results of the calculations with the STA program are the main source of
information for ANSTE program.
A NSTE Experi System
Expert systems are among the most promising and fiequently used artificial
intelligence (AI) systems [15] “We are drowning in information but starved for
knowledge. A1 will help relieve our information overload. But perhaps the most
important aspect of A1 is that it will force or encourage the conversion of information
into knowledge that can be applied to the solving of a problem or the making of a
decision. The idea is not to replace human beings but to provide us with a more
powerful tool to assist us in our work” [161.
There is extensive literature on the use of expert systems in the chemical industry.
Not many expert systems have been developed which are particularly related to the
sugar industry. During the last decade such systems have been used mainly in
industrial process operation and control [17-20], or diagnosis and prediction [2 1-22].
None of these expert systems is either related to a process simulation program or
aimed at analyzing cane sugar factories energy systems.
There are some factors that make the analysis of energy systems in cane sugar
factories become a very convenient task for the application of an expert system.
0
Energy systems in cane sugar factories constitute a clear and restricted domain.
0
The analysis has a moderate level of complexity and requires expert
knowledge.
The use of indicators of behavior allows the knowledge to be represented.
There are few experts in Cuba in comparison with the 156 sugar factories that
exist Many of these experts are going to retire in the near future, and it would
be very convenient to capture their knowledge and experience and make it
available to a wider audience.
Energy systems in cane sugar factories are very likely to change due to the
great potential for a better use of the energy available. The structure of an
expert system allows it to update such changes easily.
0
An expert system can help to overcome limitations in the analysis of energy
systems in cane sugar factories. The analysis could be done in a more rigorous
and faster way, with a systemic approach.
The ANSTE expert system is aimed at helping the specialist to detect the
problems that affect the efficiency of energy systems in cane sugar factories, and to
establish recommendations for solutions based on the calculation results of a single
running of STA [23]. The system constitutes an aid and will never be able to replace
the specialist, who will be responsible for performing the most complex tasks in the
analysis. The system has been developed, and it cannot be otherwise, not to substitute
for the experts in this field but to conserve their knowledge and experience and make
this available to others.
170
Increasing the Energy Eficiency in Cane Sugar Factories
The ANSTE program includes the following responsibilities:
To read an input file, obtained by simulation with STA.
To identify streams and modules in the information flow diagram, and to get
additional information, necessary for the calculation of indicators. This step
makes possible the analysis of any configuration and hence it provides ANSTE
with the same flexibility that STA has in the simulation.
To calculate and show the behavior of indicators.
To detect the problems in the energy system.
To recommend solutions for the problems.
To explain the inferences made in the analysis.
To write an output file with the results of the analysis.
Owing to the hierarchical nature of energy systems in cane sugar factories, besides
indicators for equipment, global indicators and indicators for plant areas are used in
the ANSTE program; all of which are essential to consider the multiple interactions
and to best characterize the energy system. Indicators for equipment are calculated in
the STA program, whereas global indicators and indicators for areas are calculated in
the ANSTE program.
All of the indicators have to be qualified. To this end, a quality is established for
each indicator when comparing it with a range or a value. The range or value, due to
its highly heuristic origin, must be established for the specific conditions of each
factory and should express what is considered desirable; for this reason the system
provides the means to adjust it. The program also allows the user to visualize the
value of each indicator and to know if they are beyond the comparison range or value.
It should be clarified that although indicators seem to bear a similarity to alarms, they
are not the same, since the ANSTE expert system up until now has been conceived for
off-line use. It is intended to be, along with the STA simulation program, a real-time
application in the future.
Once the indicators have been qualified and the additional information has been
entered, the analysis may be activated and inferential processes in the knowledge base
are performed, allowing it to diagnose problems and get recommendations for
solutions.
The identification of problems and the statement of their recommendations are
performed on the following behavior issues:
0
Surplus bagasse.
0
Use of steam.
0
Steam generation.
0
Water
0
Electricity
Installed capacities
Study of these issues, in which the most significant factors that influence the
efficiency of energy systems in cane sugar factories are included, forces the strategy
of having an integral and systemic approach in the analysis. There is significant
interaction between these issues of behavior which are taken into account.
It is considered that there is a problem related to an issue when an indicator or
some indicators are out of the comparison range or value selected. The problems
constitute factors which influence unfavorably in the specific issue. In all, ANSTE
171
0 Gozh Ledn, H Pe‘rez de Alejo and M. Rijckaerl
can detect 60 problems that affect the energy efficiency of the factory. For instance,
the most likely problems to be detected in “use of steam” issue are schematically
represented in Figure 2.
Figure 2 Problems to be detected in “use of steam” issue
The recommendations constitute the suggested actions to be taken in order to
solve problems detected in the analysis. In general, there exists at least one
recommendation related to each problem. The most time-consuming task in
developing the expert system is the gathering and organizing all of the knowledge and
experience needed to formulate these recommendations. All of the problems as well
as all of the recommendations have been described and reported [23].
The recommendations are systemic, and whenever possible, they point out
interactions (favorable or not) with other problems or indicators, both inside one issue
or among various issues. In this way the system helps to emphasize the significant
relations among the parts of the energy system and to transmit experience. For
instance, if the problem “high flue-gas temperature” is detected in the “steam
generation” issue, one recommendation among several is to control and decrease the
flue-gas temperature via the installation of heating surfaces in boilers. At the same
time, the program detects whether there is high electricity consumption in the
172
Increasing the Energy EfJiciency in Cane Sugar Factories
“electricity” issue, and then alerts that the installation of heating surfaces is going to
make the electricity consumption of the factory even higher.
In the design of ANSTE, the knowledge base was organized in 7 databases, which
contain the inferences related to problems and recommendations, with 487 production
rules in all. The knowledge represented in these bases was mainly collected through
interviews with experts and an extensive study of the literature. It also takes into
account the experience accumulated by the authors in the evaluation of many energy
systems in cane sugar factories. The ease of maintenance of the expert systems
knowledge base, allows the incorporation of knowledge and experience of further
experts, who are still being consulted to improve the system.
To create the ANSTE expert system, the KADS (Knowledge Acquisition and
Design Structuring) methodology [24] was applied and as a development tool, the
SETDC (Sistema Experto de Tablas de Decisi6n Categorial) system [25] was used.
Object oriented programming and hypertext were combined, which make the interface
easy and friendly to use. The programming of ANSTE was performed in Borland
Pascal under the DOS environment, and utilities of SETDC were used to make
inferences and develop the knowledge base. Presently a version of ANSTE
programmed in Borland Delphi and embedded in the STA program is being
implemented. Further details on the ANSTE expert system have been presented [26].
Applications
The strategy and the automated method of analysis have proven successful both in
industry and in teaching. In teaching it is an essential part of an integrated subject in
fourth year chemical engineering in two universities, and part of a subject in the
Masters Program in “Sugar Technology” in the same two universities. In industry
there are applications in most factories of four Cuban provinces, and it is intended to
use it in the rest.
An application of the automated method of analysis is now briefly presented. The
factory selected has a capacity of 6820 tonnes of cane per day and 1000 tonnes of
refined sugar per day. The thermal scheme is characterized by a high level of
complexity as shown in Figure 3. The factory delivers bagasse for the production of
board, steam to an oil factory, and electricity to the utility grid. Hence there are
multiple purposes to be satisfied by the energy system. In order to evaluate the energy
system, a “base case” was simulated with the STA program. The information flow
diagram, which consisted of 53 modules and 121 information streams, is shown in
Figures 4 and 5. A completely integrated model of the energy system in the factory
was possible.
In the information flow diagram (see Figures 4 and 5), the input information
streams are 12, 68, 75, 97, 100, 108, 128, 130, 134 and 135; the recycle information
streams are 22, 79 and 83 (all shown in large type). The modules used are as follows:
DISOT (raw melter), DISOL (dissoluter), TACH0 (refined sugar pan), CENTRI
(centrihgal), ALFIL (water heating), SECA (drier), FLAG (identifier), COOL4
(heater), EVAP67 (single evaporator), COMP2 (convergence accelerator), DIF 1 1
(subtraction), JUNCl (mixer), PAN2 (raw sugar pan station), TURBO5 (steam
turbine), SEPAl (splitter) and ENER3 (steam generator). Each block with a number is
173
0 Gozii Leon, H Perez de Alejo and M.RQckuert
a module from STA that has been assigned a specific purpose For example block 12
in Figure 4 is used to process liquor into A massecuite; blocks 34,36,37,38 and 39 in
Figure 5 are used to simulate the quadruple effect evaporator
High pressure steam
Liquor evaporator
Ouadruple died evaporator
Vaporrcll
Figure 3. Thermal scheme of the sugar factory
With the calculation modules available in the STA program it has been possible to
simulate any configuration of energy systems in all those cane sugar factories
evaluated, something that has not been possible with other software developed in
Cuba STA is completely flexible because the modules can be arranged in any order.
The lack of a graphical interface in the STA program makes the step of building the
information flow diagram tedious and time-consuming and hence not user-friendly.
This issue is going to be addressed in the new version that is being written for
Windows.
Once the “base case” was validated, the ANSTE analysis was performed. Some
calculation results obtained in the simulation of the “base case”, as well as some
indicators calculated by ANSTE, are shown in Table 3.
I 74
Increasing fhe Energy Eficiency in Cane Sugar Factories
Table 3. Calculation resultsfor the “basecase”
Shortage of bagasse
Fuel oil consumption
Factory steam Consumption
Make-up water
Specific electrical generation of the factory
Steam flow to 125/30 reducing valve
Vapor to barometric condensers in evaporators
Raw sugar pan station steam consumption
Refined sugar pan station steam consumption
Secondary heater effectiveness
Juice velocity in secondary heater
Vapor-cell evaporation rate
Turbogenerator no.2 specific steam rate
Specific steam generation
W.(n
68 6 tonnesh
13.5 m3h
85.2 YOof cane
27.6 YOof boiler feed water
52.8 kW.h/tonne cane
42.7 YOof total consumption
12.7 YOof total evaporation
22.9 YOofcane
0.86 tonnes steadtonne
refmed sugar
0.42
1.38 m / s
16 6 kg/m2.h
17 kg/kW.h
1.9 kg steam producedlkg
bagasse burned
aim
Slum Vwa
17t
t.
Figure 4. Informationflow diagram of the refined sugar section in the factory
175
0 Coza Ledn, H Pe'rez de Alejo and M.Rijckaerr
Table 4 shows a review of problems detected by ANSTE These results made it
easy to detect existing problems and their interactions. Combined with the
recommendations given, which are too extensive to be presented here, these results
were very helpful for defining solution alternatives Recommendations are related
both to operational and configuration problems
All users of ANSTE including the experts, recognize that just the presentation of
all qualified indicators is of great help for analysis of the vast and diverse information
obtained in any simulation with the STA program in a faster and standardized way.
Users, especially non-experts, also recognize that a very useful feature is its capability
to perform the analysis with a systemic approach.
Figure 5. Informationflow diagram of the raw sugar section in thefactory
I76
Increasing the Energy EfFciency in Cane Sugar Facrories
Table 4. Problems detected by ANSTE in the "base case" studied
,
Surplus Bagasse
High global steam consumption
Low steam generation efficiency
Low availability of bagasse
High amount of bagasse for other uses
Use of Steam
High process water usage and heat loss factor in pan station
High flow of secondary steam to barometric condensers
Exhaust steam to heat clarified juice
Heating of juice only in one step
High no conventional steam consumption
No atemperators at the outlet of steam turbines
Steam Generation
Low temperature of feed water to boilers
High outlet temperature of flue gases
Water
Make UD feed water to boilers
Electricity
High electricity consumption
High steam flow to reducing valve
Low turbogenerators loading
Installed Capacities
Low effectiveness in heaters
High outlet temperature differences in heaters
Low juice velocities in heaters
Low evaporation rates
High specific steam rates in turbogenerators
Low capacity usage in steam generation
Based on the problems detected and the recommendations given by ANSTE, a
study of 14 alternatives was performed [27]. These alternatives allowed the definition
of behaviour trends for the system as well as the determination of the best ways to
increase its efficiency. The alternatives included changes in the configuration of the
thermal scheme, both in the high pressure area and in the process area, as well as
changes in equipment parameters. These included heat recovery fiom condensates for
heating mixed juice, usage of vapor bleedings in first, second and third effect of the
evaporator to juice heaters and pans, variations of the grinding rate and refined sugar
production, increase of the boiler feed water temperature and of the boiler thermal
efficiency, changes of schemes in pan station, substitution of prime movers and
changes of temperature in juice streams. Among all the alternatives studied, Table 5
shows the results obtained in the economic evaluation of only five of them,
considering an economic life of 5 years and an interest rate of 15%. These alternatives
were the ones of greatest interest to the managers of the factory according to the status
of the equipment and the possibilities for investment.
177
0 God Lebn, H Perez de Alejo and M. Rijckaert
Table 5. Economic evaluation ofjhe alternatives I - clarifiedjuice heater using
secondary vapor, 6 - changing to a back pressure turbogenerator, 7- electrical
motors instead of steam machines, I0 - combination of alternatives 6 and 7, 12 combination of alternative I0 with changes in the heating-evaporationscheme.
AlterCOF
CRV
CWT
IEL
NPV
IRR
native
($/year)
($/year)
($/year)
($/year)
($)
(“/.)
Base Case 2 742 077 407 808
24910
561 600
90555
561 600
1
2719686 404352
23743
1799 168
113
6
2742077 116640
24911
993 600
7
2643 313 467296
17 556
527 040
-631 422
984046
43,4
933 120
10
2643313 196 128
17453
3030607 94,2
14239
846 720
12
2018938 125280
Notes: COF cost of fuel; CRV - opportunity cost related to high steam flow through
reducing valve; CWT - water treatment cost; IEL - cash income by electricity sale,
NPV - Net Present Value, IRR Internal Rate of Return.
-
-
It can be observed that, except for Alternative 7, the alternatives analyzed are
profitable, and that Alternative 12 is the one with the greatest benefits. In Alternative
12, fuel oil consumption is reduced from 13.5 to 10 m3/h and electricity delivered to
the utility grid is increased from 6500 to 9770 kW, representing 26% decrease and
50% increase respectively in comparison with the “base case”.
The application of the strategy and the method of analysis allowed a rigorous
characterization of the energy system to be performed, and the definition of a set of
measures and recommendations to make better use of energy. According to what
engineers in the factory stated, for the first time it was possible to perform an analysis
of the energy system integrating both the process and the steam and electricity
generation areas. These measures have given the factory managers the possibility to
ratify investments, to undertake new investments, and to set operational ways to
achieve greater cogeneration and a more rational use of bagasse and generated steam.
Conclusions
A computerized method of analysis of energy systems in cane sugar factories has
been implemented, which combines the use of process simulation and an expert
system. These techniques have been implemented with two very useful programs:
STA simulator and ANSTE expert system. The use of a sequential modular
simulation program provides an analysis method with great flexibility for the integral
evaluation of many different configurations of energy systems in cane sugar factories.
The use of an expert system provides the method with the same flexibility for the
analysis, which may be performed faster. The expert system is of great help,
especially for non-experts, to achieve systematization, unification of criteria and a
systemic approach in the analysis. The results obtained in many applications
guarantee the method of analysis as a valuable aid for the establishment of ways to
allow the increase of the energy efficiency in cane sugar factories. The use of a
sequential modular simulation program and an expert system as part of the method of
analysis makes it open for hture improvement with minor effort.
178
lncreasing the Energy Eficiency in Cane Sugar Factories
References
1
2
3
4
5
6
7
8
9
10
Albert-Thenet, J 1991 Steam economy in cane sugar factories Proceedings of the International
Conference on Energy from Sugar Cane, pp 223-243, Hawaii
International Cane Energy News 1995 Newsletter of the International Cane Energy Network, August
Kinoshita, C M I996 Sugar Factory Energy Data Protocol, Published by International Cane Energy
Network, USA
GozC 0 1996 Estrategia de analisis para lograr un us0 m&s eficiente de la energia en 10s centrales
azucareros Tesis de Doctorado, ISPJAE, C Habana
Boston, J F 1995 AspenTech is integrating steady-state and dynamic modeling The Aspen Leaf,
February, p 2
Dimian, A 1994 Use process simulation to improve plant operations Chemical Engineering Progress,
September, pp 58 - 66
Russell, L W 1994 The application of a modem process-modeling package to CSR refineries Int
Sugar J ,March, Vol96, No 1143
HYSYS Documentation Suite CD ROM, AEA Technology
Feltborg, G , R Wimmerstedt I994 Computer model and simulations on the energy situation in beet
sugar factories Zuckerindustrie, No 1, pp 30-38
Weiss, L W 1999 Cane Factory Process Modeling using SugarsTMfor Windows@ American Society
of Sugar Cane Technologists, 29 th Annual Joint Meeting of the Florida and Louisiana Divisions, June
16 18, Ft Walton Beach, Florida
Sabadi, R 1996 Aplicaciones informaticas en la industria azucarera Revista ICIDCA Sobre 10s
Derivados de la Calla de Azbcar, Vol30, No SPECIAL, pp 1-16
Perez de Alejo, H ,P Friedman 1979 Simulation of a sugar factory Int Sugar J , Vol 81, pp 67-71
Go& 0 , H Ptkez de Alejo 1999 Simulacibn modular secuencial de 10s sistemas termoenergeticosen
fabricas de azucar refino Revista de la Facultad de Ingenieria Quimica de la Universidad Autbnoma de
Yucatan, Numero 36, pp 45-55
Go& 0 , H Perez de Alejo, L Santos, J Castellanos 1996 Manual de Usuario del programa
TERMOAZUCAR (Resumen), Version 3 I, ISPJAE-UCFGOS
Crowe, E R , C A Vassiliadis 1995 Artificial Intelligence Starting to realize its practical promise
Chem Eng Prog ,January, pp 22 - 3 1
Frenzel, L E Crash course in artificial intelligence and expert systems S A M S Edicibn
fotorreproducida
Alonso, C , C Prada, J A Rico, F Hernandez 1992 An expert system for a control coordination
problem Advanced control of chemical processes (ADCHEM '91) selected papers from the IFAC
Symposium, Toulouse, France, 14-16 October 1 9 9 1 , pp 285-290
Aldini, L 1998 Utilization of fuzzy logic in diffisers Industria Saccarifera Italiana, 91(3), pp 83-87
Filippini, F , S Urbinati, M Solimano 1992 An expert system for the prevention of mahnctions and
technological optimization in a beet sugar plant Sugar y Azucar, Vol87, No 7, pp 22-23,26-27
Pilgrim, A C ,T Sewdass 1996 An expert system for the clarifier and evaporation station at a sugar
factory Proceedings of the West Indies Sugar Technologists XXV Conference, Belize, 18-21 April
1994, pp 342-35 1
Watkins, C S 1993 Condition monitoring at Illovo with special emphasis on the expert system
Nspectr I1 Proceedings of the Annual Congress South African Sugar Technologists' Association,
NO67, pp 170-175
Dohnal, M , D C Walthew 1995 The use of fuzzy expert systems to examine vague and complex
South African Sugar
problems in sugar engineering Proceedings of the Annual Congress
Technologists' Association, No 69, pp 186-190
Gozk 0 , H Perez de Alejo 1996 Manual de Usuario del Sistema Experto ANSTE, Versibn 1 0,
ISPJAE, C Habana
Hickman, F 1989 Analysis for knowledge based systems A practical guide to the KADS
methodology
Medina, J E 1995 Sistema experto de tablas de decisibn categorial Tesis de Doctorado, ISPJAE,
C Habana
God,0 , H Perez de Alejo, M Rijckaert 2000 An Expert System for the Analysis of Energy Systems
in Cane Sugar Factories The New Review of Applied Expert Systems and Emerging Technologies,
Volume 6, pp 65-80
GozP, 0 , H Perez de Alejo, M Rodriguez, B Hernhdez 1999 Mejoras de la cogeneracidn en un
central con refineria, Revista ATAC, Julio - Diciembre, pp 46-50
-
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
-
-
-
Received 30 November 2000; Accepted after revision: 15 May 2001
I
179
Документ
Категория
Без категории
Просмотров
1
Размер файла
843 Кб
Теги
expert, efficiency, cane, simulation, increase, factories, sugar, energy, system, use
1/--страниц
Пожаловаться на содержимое документа