dictum : Decision support system for analysis and synthesis of large-scale industrial systems. Part II: Databases and industrial applications

dictum : Decision support system for analysis and synthesis of large-scale industrial systems. Part II: Databases and industrial applications

Computers in Industry 18 (1992) 145-153 Elsevier 145 Applications DICTUM: Decision support system for analysis and synthesis of large-scale industr...

733KB Sizes 0 Downloads 44 Views

Computers in Industry 18 (1992) 145-153 Elsevier



DICTUM: Decision support system for analysis and synthesis of large-scale industrial systems. Part II: Databases and industrial applications L. D i e t z s c h , U . K r o e i l e r a n d K. H a r t m a n n Institute for Chemical Technology, Rudower Chaussee 5, 0.1199 Berlin, Germany

Received February 2, 1991 Revision accepted September 24, 1991 This paper presents applications of the decision support system DICTUM.The developed software package DICTUMrepresents a sophisticated, user-friendly computerized system for the analysis and synthesis of large-scale industrial process systems, such as the chemical and related industries. The generation, evaluation and optimization of alternative strategies of the development of technological systems are closely connected with computer-aided decision making. Important decision-making problems with respect to energy saving and environmental impact are characterized by the following features: (i) analysis and evaluation of the status quo, formation

of optimal control strategies and revamp policies; (ii) synthesis of feasible development strategies (long-term planning alternatives including resources requirements and energy supply, product demand) and environmental impact analysis. Some of the problems under study are the optimal combination of conversion processes in refinery plants, the improvement of alternatives for the supply of synthesis gas in an ammonia plant and the evaluation of C2-chemistry subsystems to produce acetylene, acetaldehyde and acetic acid. Keywords: Decision support system, Chemical industry, Ap-

plications, Carbon-based industry, Power production industry, Heat production industry.

1. Overview of the application fields [1,2] The chemical and related industries (oil, petrochemical, gas, carbochemical (coal-based and others) represent large-scale, complex and constantly changing industrial process systems.

T h e spectrum of products and process technologies, the product demand, the feedstocks and energy sources and their prices are subject to rapid modification. Existing processes have to be revamped and intensified, new highly efficient processes are introduced into the network of processes. Some important decision-making problems in these large-scale industrial systems are characterized in the first part of our study [3]. Primarily the decision support system (DSS) rncTUM has been developed for decision-making in carbon-based industries (power and heat production, chemical and related industries). In Fig. 1 the complex interconnections among the main subsystems of carbon-based industries are shown schematically. The DDS rncruM is applicable for any complex system described by linear i n p u t - o u t p u t relations. The present paper, part II of our study, presents details of the data banks, the different chemical industrial systems and some simple examples for application.

2. Short characteristics of the main subsystems and databases The main subsystems of the carbon-based industries are:

0166-3615/92/$05.00 © 1992 - Elsevier Science Publishers B.V. All rights reserved



Computers in Industry

ICHEMISTRY c1 / c2 I [








Fig. 1. Main industrial subsystems in DSS DICTUM.

Ludwig Dietzsch was born in 1943. He studied Chemical Engineering at the Merseburg Institute of Technology and received his Dr.-Ing. degree in Chemical Engineering in 1973. From 1970 to 1988 he worked as assistant professor of Chemical Process Systems at the same institute. Since 1988 he is a research associate at the Institute for Chemical Technology in Berlin. His special interests include the analysis and synthesis of process systems, computer-aided process engineering, decision support systems and technology assessmerit,

Klaus Hartmann is professor/chairman of Chemical Engineering at the Institute for Chemical Technology in Berlin. He received his PhD in Chemical Engineering at the St. Petersburg Institute of Technology (USSR) and his Dr.-Ing. habil, at the Merseburg Institute of Technology. He has published more than 150 articles and several books on various chemical engineering topics (analysis and synthesis of chemical process systems, fuzzy modelling, decision support systems). Uwe Kr/iller was born in 1959. He studied Chemical Engineering at the Merseburg Institute of Technology from 1980 to 1987. He received his degrees Dipl.-Ing. in 1985 and Dr.Ing. in 1988. Since 1987 he has been working at the Institute for Chemical Technology in Berlin as a scientific co-worker in the department of Chemical Engineering. His special interest is in computer-aided process engineering.

- the oil-processing and petrochemical industry; - the carbochemistry based on hard- and brown

coal and the gas industry based on natural gas (NG) and gas generation from coal and liquid carbon carriers (city-gas and synthetic natural gas (syngas-SNG); - the chemical process systems based on NG and SNG (Ci-chemistry) and the chemical process systems based on ethylene and acetylene (C 2chemistry). The following is a short description of the main characteristics of these subsystems.

2.1. The oil-processing and petrochemical industry The model of this subsystem includes various plants of crude oil processing and petrochemistry. It contains distillative separation of crude oil, processing its light fractions and conversion of the residues. The first separation takes place under atmospheric pressure. The straight run distillates produced are partially processed in refining and reforming plants and blended to products with different qualities, for example gasolines, diesel oils, jet fuel. The model also includes polymerisation, alkylation and isomerisation processes to gasoline components.

Computers in Industry

L. Dietzsch et ai. / DICTUM--Part H

The atmospheric residue is separated in a vacuum distillation. It can also be processed in a thermal cracker. The vacuum residue is sent to a visbreaking process to improve the quality of fuel oil or is processed in a delayed coker. The vacuum distillates can be processed in catalytic and hydro-catalytic crackers. In a refinery plant atmospheric distillation holds a central place. From there the gasoline and diesel oil pools are fed directly. Additionally gasoline and diesel components originate from conversion of high boiling fractions. This process is followed by the pyrolysis of liquid gas, straight run naphtha, hydro-treated diesel oils o r / a n d residue of a hydro-catalytic

cracker. The resulting olefines (ethylene, propylene) can sold. The produced C4-fraction can be processed in an extraction plant to butadiene and from the C5+-fraction aromatics (benzene, toh, ene, xylene) can be obtained. The hea~j residues of the refinery, such as residues from the vacuum distillation, from the visbreaker or from the high-conversion soaker cracker, are sent to the power station or are processed to hydrogen or methanol. The main processes included in the model are: atmospheric distillation, vacuum distillation, reforming, naphtha refining, gasoil refining, hydrocracking, catalytic cracking, thermical cracking, high-conversion soaker cracking, delayed coking,


' I

I olefines,

I i-butane



pol erisatio


I >1 alkylation


'>Iisomerisation _~_u.




crude>"'~l~~'= >>~----~ oil ref~i--~>'I-" ning ~ ~'<-~~ _ ~extaromat r___.[ .actico.s__~n ~


I refining I I cracked naphtha I


I atm. residue


propylene V




Ibutadiene ~> butadiene extraction

,> gasoline > gasoil

.>. ~






> fuel oil

Fig. 2. Simplifiedstructure of crude oil processingsubsystem.



Computers in Industry

electric power power supply

heat >

lignite feed city gas



hard coal feed i


city gas and synthesis gas supply

synthesis gas coke

natural gas feed



fuel oil liquid products processing



Fig. 3. The main inputs and outputs of the carbochemicai and gas industry.

visbreaking, polymerisation, isomerisation, alkylation, pyrolysis, butadiene extraction, extraction of aromatics, methanol synthesis, MTBE-synthesis, oil power station, gas power station, steam reforming of natural gas, steam reforming of refinery gases, and partially oxidation of oil. Figure 2 shows the main processes included in the model of crude oil processing and petrochemistry and their simplified structure.

2.2. Carbochemistry and gas industry In some coal-rich countries the gas industry is closely connected with the carbochemical industry, including power generation. Figure 3 gives an overview over the feedstocks and main products of the whole carbochemistry. Hard coal and lignite together with different sorts of natural gas are the main feedstocks of this system. The main products of this system are cokes, various liquid products and gases (city gas and synthesis gas). The city gas is produced by the following technologies: - coal gasification (high pressure, fixed bed);

carbonization of lignite (high temperature); - coking of hard coal; decomposition of natural gas (high pressure); gasification of pulverized lignite (GSP process); gasification of liquid products (tar, fuel oil); decomposition of liquid gas (low pressure, catalytic). The by-products of these technologies are tars, fuel oil, middle and light oils, phenols. They are processed by the chemical industry for producing green coke, naphtalene, pyridine, paraffins, components of gasoline and lubricants. A simplified structure is shown in Fig. 4. -





2.3. Subsystems of the natural gas/syngas and ethylene/acetylene industries (C l- and C2-chemistry) The subsystem of the C~-chemistry concerns processes which are based on natural gas and synthetic natural gas (syngas) as feedstocks. Power- and heat-production processes based on gas are not included in this subsystem. They belong to the "energetics" subsystem.

Computers in Industry

L. Dietzsch et aL / DictuM--Part H


[briquet ~.--1 ignite-~ lignite~I ignite--.4power/heating[ plant | l mining| [station




i______~ ,, coal feed liquidgas ", "=='I feed

riquettes --"'






Vm Vm


ldecomposit.[coal [h.-coaX-[xignitJ [uecomposXt. [gasifier look,~ [carboniz.[liquidgas












~ _ .

natural gas v --cityfeedgas ~~i~d~~



. ig,,i~,~ I l

] V ~m~h~l~e

Fig.4. Simplified

[oil l tar


scheme of carbochemistry and city gas production.

A simplified structure of the C~-chcmistry processing system is shown in Fig. 5; it includes such important products as methanol, ammonia, olefins and aromatics. These key products are feedstocks for the "methanol chemistry", which is represented in a simplified form in Fig. 6. The C2-chemistry subsystem includes the processes of the traditional acetylene and ethylene

routes as well as new highly efficient processes based on these feedstocks. This subsystem consists of about 100 processes with the following substructure: - the acetylene and ethylene production processes (a simplified structure is shown in Fig. 7); - the so called primary processes for the production of important intermediates, such as ac-


methanol, [

ethylene-I< glycol


organic [ dixso- <

! i








o omotiool


I o= oo.o.I aim-mlv[ oinziearusr,eafer!t< ' ~ Fig.5. Simplified

structure of processes based on synthetic natural




Computers in Industry

I acetic


oxalic acid


I chlormethane ~-- ~ methanol

methyl acetate I V ethylacetate acetic acid anhyd. ethylidendiacetate




aliphatic amlnes

--,{ ooetonitrilo


formaldehydev[ I <

olefines aromatics gasoline


aldehyde UF-resins



Fig.6. Subsystemof the methanolchemistry. etaldehyde, acetic acid, ethylene oxide and others; the consecutive processes based on acetaldehyde, acetic acid and other important products (see Fig. 8 for an example of this conversion tree).

3. Selected applications of the DSS DICTUM

The DSS DICTUM is widely used for quite different decision processes on various levels [4,5]. In the following some simple examples of application are described.

3.1. Analysis and improvement of the efficiency of an oil-processing system In a refinery plant with the structure shown in Fig. 9 two types of crude oils are available for processing, Arabian light and Forties. The existing plant includes an atmospheric distillation, a thermic cracker and blending stations for gasoline and fuel oils. By investment this structure can be completed by a vacuum distillation and a hydro-cracker. These plants should be integrated in the existing structure shown in Fig. 10. The profit should be maximized. The results of the solution of this problem are

Ifuel1 gas I

coke lime



~ > vinyl chloride tetrachlormethane




>~calcium-~carbid_>t acetylene tar... >~process I Electricity->i , / [acetylene v '

by! products[




lime nitrogen


1,4 - butindiol

l- DC arc

1~ plasma •




acetylene black

Fig, 7. Simplifiedstructureof the acetylenesubsystem.

Computers in Industry

L. Dietzsch et al. / DiCTUM--Part H



,, • aldehyde

lacetylene I I

• butanol

v I



• hexanol

/ Idi°l-{:'3}

• butan-





o o n o

acetaldol I

•lphthalate I

->Idibutyl- [ lPhthalate

acetv !




--> hexantriol --[>[butadiene

Fig. 8. Subsystem of the consecutive products of acetaldehyde. shown in Table 1 as variant 1. All options (both types of crude oils and investment capital) are used. Including also suboptimal solutions in this application, it is possible to construct the following variants: Variant 2: both types of crude oils but no investment capital available; Variant 3: only Forties and no investment capital available. Variant 4: only Forties and investment capital available. Variant 5: only Arabian light and investment capital available. The results of these variants are also shown in

Table 1 and indicate that Forties is more suitable for hydro-cracking than Arabian light. In the example no product demand or oil supply are formulated. The capacities of the atmospheric distillation and of the thermic cracker were limited. The model was driven only by economic criteria.

3.2. Improvement of alternatives for the supply of synthesis gas in an ammonia plant The production of ammonia in a chemical plant is to be expanded. With respect to this it is necessary to enlarge the production of synthesis

>I blending F > gasoline

[ I naphtha



: .

crude • • oil , kerosine ] • - • dist.| [ --• •. }-gasoil-J-•~ blending h Arabian I - I gas.1 , .

., .I I ,o .H arm. cker ' I


V V t



fuel oil (light)

I ttom 1 i

V ,,

naphtha kerosine



> fuel oil (heavy)

Fig. 9. Refinery structure under s t u d y .

Computers in Industry 152


m----->1 blending ~ > gasoline vT--->>naphtha crudem ]-,}; kerosine oil | kerosine--l--• fuel oil dist'~asoil [-- -->I blending 1 (Iight) I naphtha





, il It, r.! ~l-•Icra-| I II Ibottom I





I hyd og .





" r>l r ke l Fig. 10. Integrated structure of the refinery.

gas--hydrogen--too. Three alternatives for producing ammonia are considered: Variant 1: (3SP process (gasification of pulverized lignite); Variant 2: steam reforming; Variant 3: a modernized Winkler process. These technologies and the ammonia production are the main processes in the alternative systems.

Furthermore, the systems include new technologies for heat and energy supply, lignite mining and mending, LURGI carbonization, transport of natural gas and so on. The variants were evaluated by the following criteria: costs, investments and expense of primary energy. The results of evaluation are shown in Table 2. It is obvious that the variant 2--the ammonia

Table 1 Selected results of optimized variants of crude oil processing Variant

(in 10~'$) Profit Proceeds Costs Investment (in 106 t) Gasoline Naphtha Fuel oil heavy Fuel oil light Kerosin Arabian light Forties






151.51 2675.94 2524.43 429.43

90.48 1563.49 1473.01 -

17.08 1562.83 ! 545.75 -

97.02 2729.78 2632.76 490.95

135.94 1694.79 1558.85 248.11

2.92 2.79 5.11 5.50 2.66 11.90 7.39

2.13 3.66 4.23 3.33 1.55 11.90 0.0

1.74 3.90 5.43 5.87 3.00 20.00

2.57 1.20 2.95 3.33 1.55 11.90 -

0.80 1.20 4.91 3.67~ 1.88 12.50

L. Dietzsch et al. / otcruM--Part H

Computers in Industry Table 2 Evaluation of variants for ammonia production Variant Costs Investments Expense of primary energy




1.0 1.0 1.0

0.95 0.88 0.81

1.27 0.84 3.12

plant on the base of steamreforming--is favoured. However, the costs by the steam reforming variant is equivalent to the GSP process variant for a 6.7 percent higher price of natural gas. In view of a decision support it is possible and necessary to include further criteria for an evaluation, for instance the pollution of the environment or the valuation of by-products.


When comparing the alternative routes 2 and 3 with the reference route the following conclusions may be made: (a) The production costs increase generally, the material and energy costs increase only in route 2; (b) The consumption of electric energy decreases in both alternative routes, significantly so in route 3; (c) Essential improvements are attainable through the alternative routes regarding transport expenses, emissions and man powers. These simple examples show how DICTUMallows studies and decision making in technological structures in an interactive dialog. DICTUM also can be used for integrating marketing and operations with short-term tasks.

3.3. Evaluation of alternatives for the C2-chernis,r2 subsystem The fundamental technological structure of the different routes to produce acetylene, acetaldehyde and acetic acid is very complex. As an example three alternative routes have been investigated: Route 1: traditional acetylene route (reference route); Route 2: production of acetic acid by a new process based on SNG and methanol and of acetylene partially by plasma as well; Route 3: new processes to produce acetylene (by plasma), acetaldehyde (from athylene), and acetic acid. The results of evaluation are shown in Table 3 (only selected parameters). Table 3 Comparison of the results of the three alternative routes Routes Raw materials and energy costs Proceeds Electric energy Production costs Transport expense Emissions




1.00 1.00 1.00 1.00 1,00 1.00

1.09 1.15 0.79 1.03 0.45 0.49

0.86 1.16 0.42 1.08 0.01 0.001

References [1] A.-M. Barnikow, U. Behrendt, K. Hartmann and M. Scharni, "An interactive decision support system based on multicriteria optimization for application in energy and chemistry sectors", IFIP System Modelling and Optimization, Leipzig, Germany, July 1989. [2] A.-M. Barnikow, U. Behrendt, K. Hartmann and M. Scharni, "Decision support system for large-scale development planning in chemical industry", Proc. 3rd Int. Conf. on Artificial Intelligence in CIM, Leningrad, USSR, April 16-19, 1990. [3] A.-M. Barnikow, U. Behrendt, K. Hartmann and M. Scharni, "DIC'rtJM: Decision support system for analysis and synthesis of large-scale industrial systems. Part I: Components," Computers in Industry, Vol. 18, No. 2, 1992, pp. 135-144. [4] L. Dietzsch, K. Hartmann and U. Behrendt, "A computer-aided decision support system for the optimal long-term planning in large-scale chemical and related process systems", in: H.Th. Bussemaker and P.D. ledema (eds.), Computer Applications in Chemical Engineering, (Proc. Eur. Symp, ComChem '90, The Hague, The Netherlands, May 1990), Elsevier, Amsterdam, 1990. [5] L. Dietzsch, K. Hartmann, U. Behrendt and U. Kroeiler "A decision support system for the development of the carbon-based process industries", lOth Congres CHISA '90, Prague, CSFR, August 26-31, 1990.