Evaluation of thermodynamic models for pure hydrogen and hydrogen-containing streams

Charalampidi, Antonia Maria; Χαραλαμπίδη, Αντωνία Μαρία

dc.contributor.author	Charalampidi, Antonia Maria	en
dc.contributor.author	Χαραλαμπίδη, Αντωνία Μαρία	el
dc.date.accessioned	2022-12-05T09:47:31Z
dc.date.available	2022-12-05T09:47:31Z
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/56347
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.24045
dc.rights	Default License
dc.subject	Data evaluation for hydrogen-containing systems	en
dc.subject	Thermodynamic modeling of pure hydrogen	en
dc.subject	Thermodynamic modeling of hydrogen-containing binary systems	en
dc.subject	GERG- 2008 EoS evaluation of hydrogen-containing systems	en
dc.title	Evaluation of thermodynamic models for pure hydrogen and hydrogen-containing streams	en
heal.type	bachelorThesis
heal.classification	Θερμοδυναμική	el
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2022-07-11
heal.abstract	Since the world is moving towards a new era which is based on zero greenhouse gas emissions in the atmosphere and renewable energy resources, emphasis should be placed on a substitute solution for the fossil fuels; hydrogen. Pure hydrogen, which is mainly produced by methane steam reforming followed by carbon dioxide capture and secondarily by water electrolysis, can be the “green weapon” against today’s environmental crisis. Nevertheless, in order to achieve the transition from fossil fuels to hydrogen, it is important to test if the existing pipeline and storage system that is used for natural gas streams can be applicable to hydrogen-containing streams as well. In this work the main focus is paid to the properties related to pipeline design and storingconditions determination for the cases of pure hydrogen and of binary mixtures between hydrogen and basic natural-gas components. More specifically, several thermodynamic models have been evaluated in terms of prediction-accuracy comparing to the available online experimental datasets of the desired examined properties. The properties of major importance as it comes to transporting and storage are; vapor pressures, as it comes to pure components, or vapor-liquid equilibrium information, as it comes to mixtures, which can be used to properly determine the operating conditions of the production chain for the desired mixture-phases, the system’s density, which is important for calculating the frictional pressure loss inside a pipe, the energetic properties of molar heat capacity and enthalpy and also Joule-Thomson coefficient, which are important for determining temperature changes of a fluid, and lastly the speed of sound, which is important for defining the critical mass flux of pipeline flows. Additionally, the components that were considered as components of interest and were examined when mixed with hydrogen in terms of binary mixture were the main components of natural gas stream that can be found in hydrogen’s production via steam reforming and these are: methane, which is the most important one, ethane, propane, carbon monoxide, carbon dioxide and nitrogen. Lastly, the thermodynamic models that have been tested for the abovementioned calculations are; GERG-2008 EoS which is the reference equation for natural gas streams and similar gases’ streams and it is important to see if it can perform high accuracy for mixtures of components of these streams when mixed with large quantities of hydrogen, the classic cubic equations of state Peng-Robinson and Soave-Redlich-Kwong which are very simple to use and are reliable for various mixtures regardless their simplicity, a more complicated statistical model such as Perturbed-Chain SAFT equation of state and lastly the UMR-PRU predictive model which is the PR EoS coupled with UNIFAC through the Universal Mixing Rule. After a detailed searched in the available online literature, various experimental data points referring the abovementioned systems of pure hydrogen and hydrogen-containing binary mixtures were collected. The data available for pure hydrogen were plenty, they were reliable and lead to safe conclusions as it comes to the model accuracy comparison. This was not the case for the data available for the binary mixtures. The thermophysical properties’ data have been extracted mainly from rather old sources, since there were no up-to-date data found, and have a clear focus on single phase density values for the systems of hydrogen mixed with methane, carbon dioxide or nitrogen for very low concentration of hydrogen in the mixtures. Based on such data, it is not possible to draw safe conclusions regarding modern processes iii related to hydrogen production, since these will be based on high hydrogen concentrations. The available vapor-liquid equilibrium data led to safer results. From the thermodynamic model comparison, it was concluded that the GERG-2008 EoS can be used as a reference equation for pure hydrogen since it gave the best results in all of the calculations. GERG-2008 EoS can also be trusted for the single-phase density and the sound velocity calculations of the six binary mixtures regarding the experimental data. Since the data fail to cover a wide range of hydrogen compositions, GERG-2008 EoS was used to extend the existing database for single-phase density and speed of sound data in higher hydrogen compositions and it was found that the rest of the thermodynamic models give similar results for a composition range from 0 % to 100 % hydrogen, with SRK EoS being the most reliable comparing to the GERG-2008 EoS. Both in case of pure hydrogen and hydrogen-containing binary mixtures, no model can predict well enough the residual part of heat capacities or the Joule-Thomson coefficient. The vapor-liquid equilibrium calculations were better predicted by UMR-PRU EoS. It is a fact now that for these six binary mixtures the models that can predict satisfactorily their thermophysical properties do not succeed on vapor-liquid equilibrium and vice versa. Due to this fact, the optimum binary interaction parameters of the mixtures for, indicatively, PR EoS were properly determined, after carrying out fitting to the bubble point pressure experimental data in order to realize if this technique can lead the equations of state to achieve better accuracy for both vapor-liquid equilibrium and other properties calculations. Finally, it was concluded that the reference equation, GERG-2008, can be trusted when predicting pure hydrogen’s properties since it’s the most accurate of the evaluated models, it is the most accurate model as it comes to the prediction of thermophysical properties of the binary mixtures that contain hydrogen, which was expected since the model’s parameters have been determined after a fit to the experimental data, but it does not give so good results regarding the vapor-liquid equilibrium for the datasets that have been examined in this thesis. It should be examined further weather a more proper determination of the parameters of the model’s mixing rules can lead to better results, with a focus on better accuracy regarding the vapor-liquid equilibrium calculations. As it comes to both pure hydrogen and hydrogencontaining binary mixtures and the thermophysical properties predictions, PR EoS achieves better accuracy when it is paired with hydrogen’s acentric factor that is not its experimental but a fitted one while at the same time SRK EoS behaves better with the experimental acentric factor for hydrogen. In terms of vapor-liquid equilibrium predictions, both PR and SRK EoS behave better when hydrogen’s experimental acentric factor is applied. A proper determination of the binary interaction parameters can lead to significant improvements regarding the VLE predictions. UMR-PRU model shows a very good accuracy when predicting the vapor-liquid equilibrium of the mixtures while at the same time it does not result in the lowest deviations regarding the other properties of the same mixtures. This behavior is expected since the model’s parameters were determined after regressions based on vaporliquid equilibrium experimental data. Lastly, PC-SAFT EoS cannot be trusted for calculations related to hydrogen-containing streams while in this case better determination of the binary interaction parameters could also lead to significant improvements of the final results.	en
heal.advisorName	Voutsas, Epaminondas	en
heal.advisorName	Koulocheris, Vassilis	en
heal.advisorName	Tasios, Akis	en
heal.committeeMemberName	Magoulas, Konstantinos	en
heal.committeeMemberName	Lyberatos, Gerasimos	en
heal.committeeMemberName	Voutsas, Epaminondas	en
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Χημικών Μηχανικών. Τομέας Ανάλυσης, Σχεδιασμού και Ανάπτυξης Διεργασιών και Συστημάτων (ΙΙ). Εργαστήριο Θερμοδυναμικής και Φαινομένων Μεταφοράς	el
heal.academicPublisherID	ntua
heal.numberOfPages	131 σ.	el
heal.fullTextAvailability	false