Steam Cracking: Chemical Engineering
Total Page:16
File Type:pdf, Size:1020Kb
Steam Cracking: Kinetics and Feed Characterisation João Pedro Vilhena de Freitas Moreira Thesis to obtain the Master of Science Degree in Chemical Engineering Supervisors: Professor Doctor Henrique Aníbal Santos de Matos Doctor Štepánˇ Špatenka Examination Committee Chairperson: Professor Doctor Carlos Manuel Faria de Barros Henriques Supervisor: Professor Doctor Henrique Aníbal Santos de Matos Member of the Committee: Specialist Engineer André Alexandre Bravo Ferreira Vilelas November 2015 ii The roots of education are bitter, but the fruit is sweet. – Aristotle All I am I owe to my mother. – George Washington iii iv Acknowledgments To begin with, my deepest thanks to Professor Carla Pinheiro, Professor Henrique Matos and Pro- fessor Costas Pantelides for allowing me to take this internship at Process Systems Enterprise Ltd., London, a seven-month truly worthy experience for both my professional and personal life which I will certainly never forget. I would also like to thank my PSE and IST supervisors, who help me to go through this final journey as a Chemical Engineering student. To Stˇ epˇ an´ and Sreekumar from PSE, thank you so much for your patience, for helping and encouraging me to always keep a positive attitude, even when harder problems arose. To Prof. Henrique who always showed availability to answer my questions and to meet in person whenever possible. Gostaria tambem´ de agradecer aos meus colegas de casa e de curso Andre,´ Frederico, Joana e Miguel, com quem partilhei casa. Foi uma experienciaˆ inesquec´ıvel que atravessamos´ juntos e cer- tamente que a vossa presenc¸a diaria´ apos´ cada dia de trabalho ajudou imenso a aliviar as saudades de casa. As` minhas amigas Helena e Mariana da FEUP, um sentido obrigado pela vossa amizade e companhia. Espero voltar a ver-vos em breve! Giovanni e Gabriele, devo anche dire che sono orgoglioso di voi avere come amici e sono sicuro che rimarremo in contatto. Grazie mille per tutto! 병8, ¬` 과 예|, 여러분D 만나서 정Ð 반가웠고 여러분과 \l가 수 있´서 영광tÈ습È다. 영m, 포t,갈, 9@ \mÐ서 우¬가 - 다시만날 수 있È<t 좋겠습니다. 감¬i니다! To all my other friends Artur, Francisco, Maria, Mariana, Ometha, Pierre, Renato and Tomasz, it was a pleasure to share some of the best moments I had in London with you. Nao˜ poderia deixar tambem´ de agradecer a` minha fam´ılia que sempre me apoiou e continua a apoiar. Aos meus avos,´ que tenho a certeza que continuarao˜ a olhar por mim. Ao meu pai, pelos bons momentos que passamos´ e que continuaremos a ter enquanto pai e filho. Por fim, a` minha querida mae,˜ pelo amor incondicional e por fazer de mim nao˜ so´ o aluno e o homem que sou mas certamente o engenheiro que serei amanha.˜ Por ultimo´ mas definitivamente nao˜ menos importante, a` Mariana, minha namorada, pelos quatro maravilhosos anos que passamos´ juntos e pela persistenciaˆ em manter vivo a nossa relac¸ao˜ nos sete meses em que estivemos separados. v vi Resumo A produc¸ao˜ de etileno e propileno a partir de nafta e alimentac¸oes˜ gasosas como etano, propano e outros alcanos leves atraves´ do craqueamento termico´ e´ um pilar da industria´ qu´ımica. No presente trabalho, modelou-se uma fornalha de craqueamento termico´ e foram implementados e validados varios´ esquemas cineticos´ da literatura, moleculares e radicalares, usando dados de fornalhas industriais al- imentadas a etano, propano e nafta. Para as alimentac¸oes˜ gasosas, os resultados parecem sugerir que as cineticas´ implementadas sao˜ capazes de prever com exatidao˜ a distribuic¸ao˜ dos produtos a` sa´ıda da fornalha, com especial foco na cinetica´ radicalar. Relativamente a` alimentac¸ao˜ de nafta, as cineticas´ radicalares implementadas foram incapazes de produzir resultados proximos´ dos industriais. A substituic¸ao˜ do vapor de agua´ por outros diluentes foi estudada em estado estacionario,´ tendo-se con- clu´ıdo que se a temperatura de sa´ıda das serpentinas nao˜ puder ser aumentada, nenhuma diferenc¸a podera´ advir da utilizac¸ao˜ de diferentes diluentes. Contudo, se tal constrangimento puder ser alargado, o helio´ parece impor-se como a melhor alternativa. Por fim, tendo em conta que a implementac¸ao˜ de esquemas cineticos´ requer uma composic¸ao˜ molecular da alimentac¸ao˜ e que estas ultimas´ sao˜ normal- mente caracterizadas por outros ´ındices, foi desenvolvido um modelo de caracterisac¸ao˜ de alimentac¸ao˜ l´ıquida. Deste modo, poder-se-iam obter composic¸oes˜ moleculares recebendo como dados de entrada os ´ındices comerciais que habitualmente caracterizam estas frac¸oes˜ petrol´ıferas. Contudo, o modelo mostrou-se insuficiente para prever corretamente tais composic¸oes,˜ tendo-se conclu´ıdo que se teria de incluir a priori alguma informac¸ao˜ por forma a melhorar as previsoes˜ do modelo. Palavras-chave: Craqueamento termico,´ Etano, Propano, Nafta, Cinetica,´ Caracterizac¸ao˜ de alimentac¸oes˜ l´ıquidas vii viii Abstract The production of ethylene and propylene from naphtha and gaseous feedstocks such as ethane, propane and other light alkanes via thermal cracking is a cornerstone of the chemical industry. In the present work a mathematical steam cracking furnace model is presented and several kinetic schemes from literature, both molecular and radical, were implemented and validated against data from industrial ethane, propane and naphtha feedstocks processing furnaces. The results showed that, for gaseous feedstocks, the implemented kinetics were able to accurately predict product yields, with the radical scheme superseding the molecular one. Regarding naphtha cracking, however, the implemented rad- ical kinetics from literature seemed to fail at predicting plant data. A steady-state study on alternative diluents relatively to steam was also carried out and it was concluded that there may actually be no difference between diluents if one is not willing to further increase the coil outlet temperature, although helium posed the best alternative if no constraints on temperature exist. At last, since the implementa- tion of kinetic schemes requires the molecular composition of the feed and because liquid feedstocks are usually characterised by other indices rather than a detailed hydrocarbon analysis, a feed charac- terisation model was developed. This model had the objective to determine the molecular composition of naphtha feedstocks given the commercial indices that usually characterise such petroleum fractions. The results, however, showed that the model is not able to accurately determine such compositions, having been concluded that a priori knowledge had to be included to improve its predictions. Keywords: Steam cracking, Ethane, Propane, Naphtha, Kinetics, Feed characterisation ix x Contents Acknowledgments...........................................v Resumo................................................. vii Abstract................................................. ix List of Tables.............................................. xvi List of Figures............................................. xix Nomenclature.............................................. xxvi Glossary................................................ xxviii 1 Introduction 1 1.1 Motivation.............................................1 1.2 Scope...............................................2 1.3 State-of-the-art..........................................2 1.4 Outline...............................................2 2 Background 5 2.1 Ethylene market.........................................5 2.2 Steam cracking process.....................................9 2.2.1 Process description...................................9 2.2.2 Furnace.......................................... 10 2.2.3 Recovery section..................................... 14 2.2.4 Hydrocarbon fractionation section........................... 15 2.3 Steam cracking reactions.................................... 19 2.3.1 Thermodynamics..................................... 19 2.3.2 Mechanisms....................................... 20 2.3.3 Kinetic models...................................... 23 3 Implementation 31 3.1 The gPROMS® platform..................................... 31 3.1.1 gPROMS ProcessBuilder® ............................... 31 3.2 Foreign Objects.......................................... 32 3.3 Other gPROMS® tools...................................... 32 3.3.1 Optimisation....................................... 32 xi 3.3.2 Parameter estimation.................................. 32 3.4 Physical properties........................................ 33 3.4.1 Infochem MultiflashTM .................................. 33 3.4.2 gSAFT® ......................................... 34 3.5 Implementation of Large Scale Kinetic Mechanisms..................... 34 3.5.1 Sparse matrix compression scheme.......................... 35 3.5.2 Application of the LSKM foreign object......................... 35 4 Steam cracking furnace 39 4.1 Model equations......................................... 39 4.1.1 Tube model........................................ 39 4.1.2 Energy input model................................... 43 4.1.3 Cooling jacket model................................... 48 4.1.4 Furnace and coil model................................. 48 4.2 Ethane cracking......................................... 50 4.2.1 Kinetics.......................................... 50 4.2.2 Industrial case...................................... 50 4.3 Propane cracking......................................... 54 4.3.1 Kinetics.......................................... 54 4.3.2 Industrial case...................................... 54 4.4 Naphtha cracking........................................