doktorska disertacija
Klavdija Zirngast (Author), Zorka Novak-Pintarič (Mentor), Zdravko Kravanja (Co-mentor)

Abstract

V doktorski disertaciji je prikazan razvoj robustnih računalniških metod za načrtovanje in sintezo fleksibilnih procesov in mrež z velikim številom negotovih parametrov. Uporaba natančnejših metod, kot je npr. Gaussova integracijska metoda, vodi do eksponentne rasti matematičnega problema glede na število negotovih parametrov. Glavni dosežek disertacije je metodologija, s katero se izognemo eksponentni rasti. Metodologija temelji na dvostopenjski stohastični formulaciji z rekurzom in razstavi reševanje na več korakov, v katerih ločeno določimo prvostopenjske spremenljivke, tj. topologijo in velikost oz. kapaciteto procesa, ter drugostopenjske spremenljivke, tj. obratovalne in regulacijske spremenljivke. Pri tem praviloma rešujemo matematični problem le v eni točki (scenariju), v posameznih variantah metode pa hkrati v manjšem številu scenarijev, npr. do deset. %Osnovna ideja metodologije je naslednja: začetno optimalno, a praktično nefleksibilno procesno shemo generiramo pri nominalnih vrednostih negotovih parametrov. To shemo nato zaporedoma optimiramo pri različnih skrajnih vrednostih negotovih parametrov, tako da za procesne enote, ki so že v shemi, določimo potrebno povečanje velikosti, nove enote pa dodajamo le, če je to potrebno za doseganje dopustne rešitve. Na ta način določimo izbor procesnih enot in njihove velikosti za fleksibilno obratovanje. Ko se ti ne spreminjajo več, izračunamo indeks fleksibilnosti za dobljeno rešitev in izvedemo stohastično optimizacijo Monte Carlo za določitev optimalnih drugostopenjskih spremenljivk. Obenem izračunamo pričakovano vrednost optimizacijskega kriterija z določeno stopnjo zaupanja. Za vzpostavitev toka informacij med ločenima korakoma določanja prvo- in drugostopenjskih spremenljivk smo izdelali modificirano metodo, v kateri izračunavamo korekcijske faktorje, s katerimi izboljšamo vzpostavljanje kompromisov med obema vrstama spremenljivk in iterativno izboljšujemo končni rezultat.%Za povečanje učinkovitosti optimizacije Monte Carlo v zgoraj opisani metodologiji smo razvili indikator, s katerim določimo minimalno potrebno število scenarijev, da so rezultati dovolj točni za praktično uporabo. Na ta način skrajšamo čas reševanja. Vpeljali smo tudi relativni indeks optimalnosti, s katerim primerjamo približne pristope, ki smo jih razvili, z bolj točnimi.%S predlagano metodologijo smo izvedli sinteze fleksibilnih omrežij toplotnih prenosnikov in dobavnega omrežja za proizvodnjo električne energije iz bioplina, ki smo jo nadgradili s predelavo digestata v kvalitetnejša gnojila. Dokazali smo, da lahko s to metodologijo generiramo fleksibilne rešitve za velike procesne sheme z več deset negotovimi parametri v zmernem času z obvladljivim računalniškim naporom.%V zadnjem delu disertacije smo oblikovali pristope za vključevanje negotove vrednosti davka na emisije CO2 v sintezo fleksibilnih procesov v celotnem življenjskem ciklu. Razvili smo enoperiodni in večperiodni stohastični pristop. Primerjava rezultatov z determinističnim pristopom je potrdila prednost stohastičnega pristopa. %Razvita metodologija predstavlja orodje za sprejemanje trajnostnih investicijskih odločitev v pogojih negotovosti in prispeva k dolgoročnemu povečanju učinkovitosti in konkurenčnosti v procesni industriji. Njena glavna prednost je, da je uporabna za reševanje primerov z velikim številom negotovih parametrov. V nekaterih študijskih primerih smo uporabili trajnostno namensko funkcijo in tako sintezo fleksibilnih procesov povezali s trajnostnim razvojem, pri čemer se vzpostavljajo dolgoročni optimalni kompromisi med fleksibilnostjo obratovanja ter ekonomskimi, okoljskimi in socialnimi vidiki. S predelavo odpadka iz bioplinarne v koristne produkte smo v sintezo fleksibilnih procesov in mrež uvedli zapiranje zank in krožno gospodarstvo.

Keywords

negotovost;fleksibilnost;stohastično optimiranje;sinteza procesov;dobavno omrežje;emisije CO2;trajnostni razvoj;doktorske disertacije;

Data

Language: Slovenian
Year of publishing:
Typology: 2.08 - Doctoral Dissertation
Organization: UM FKKT - Faculty of Chemistry and Chemical Engineering
Publisher: [K. Zirngast]
UDC: 502.131.1:[519.85+66.011](043.3)
COBISS: 75151107 Link will open in a new window
Views: 480
Downloads: 41
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Synthesis of flexible and sustainable (bio)chemical processes and networks under uncertainty
Secondary abstract: The doctoral thesis presents the development of robust computational methods for the design and synthesis of flexible processes and networks with large numbers of uncertain parameters. The use of more precise methods, such as the Gaussian integration method, leads to an exponential growth of the mathematical problem with the number of uncertain parameters. The main achievement of this work is the methodology we developed to avoid exponential growth. The methodology is based on a two-stage stochastic problem with recourse and divides the solution into several steps, where the first-stage variables, i.e., topology and process size or capacity, are determined first, followed by the determination of the second-stage variables, i.e., operating and control variables. As a rule, the mathematical problem is solved in only one point (scenario), but in some variants of the method simultaneously in a small number of scenarios, for example, up to ten.%The basic idea of the methodology is the following: an initial optimal, but practically inflexible process flowsheet is generated at nominal values of uncertain parameters. This flowsheet is then successively optimized at various extreme values of the uncertain parameters, so that the necessary enlargement is determined for process units already included in the flowsheet, while new units are added only when necessary to achieve a feasible solution. In this way, we determine the process topology and process unit sizes for flexible operation. When these no longer change, the flexibility index for the obtained solution is determined and a stochastic Monte Carlo simulation is performed to determine the optimal second-stage variables. At the same time, the expected value of the optimization criterion is calculated with a certain confidence level. In order to establish the flow of information between the separate steps of determining the first- and second-stage variables, a modified method was developed in which the correction factors are calculated to improve the trade-offs between the two types of variables and iteratively improve the final result.%In order to increase the efficiency of the Monte Carlo simulation in the methodology described above, an indicator has been developed to determine the minimum number of scenarios required so that the results are sufficiently accurate for practical application. This reduces the computation time. In addition, a relative optimality index was introduced to compare the developed approximate approaches with the more accurate ones.%Using the proposed methodology, syntheses of flexible networks of heat exchangers and supply network for the production of electricity from biogas upgraded by processing digestate into higher value fertilizers were carried out. It was proved that flexible solutions for large process schemes with large numbers of uncertain parameters can be generated by the proposed methodology in moderate time with manageable computational effort.%In the final part of the thesis, the approaches were developed to include the uncertain value of the CO2 emission tax in the synthesis of the flexible processes during the life cycle. A single-period and a multi-period stochastic approach were tested. A comparison of the results with those of the deterministic method confirmed the advantage of the stochastic approach.%The developed methodology represents a tool for sustainable investment decisions under uncertainty and contributes to the long-term increase of efficiency and competitiveness in the process industry. Its main advantage is that it is suitable for solving problems with large numbers of uncertain parameters. In some case studies, a sustainable objective function was used that combines the synthesis of flexible processes with sustainable development and identifies optimal long-term trade-offs between operational flexibility and economic, environmental and social aspects. By processing waste from biogas plants into useful products, a closed loop
Secondary keywords: mathematical programming;uncertainty;flexibility;stochastic optimization;process synthesis;supply network;CO2 emissions;sustainable development;Matematično programiranje;Univerzitetna in visokošolska dela;
Type (COBISS): Doctoral dissertation
Thesis comment: Univ. v Mariboru, Fak. za kemijo in kemijsko tehnologijo
Pages: XVIII, 108 str.
ID: 12633406