Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Ανάλυση ευαισθησίας και έλεγχος ποιότητας εργαστηριακών μετρήσεων της μελέτης διαφορικής εξάτμισης ρευστών πετρελαϊκών ταμιευτήρων

Psarras Dimitrios

Full record


URI: http://purl.tuc.gr/dl/dias/CCF8A25C-AA58-4656-BC8B-C758CD30A69B
Year 2021
Type of Item Diploma Work
License
Details
Bibliographic Citation Δημήτριος Ψαρράς, "Ανάλυση ευαισθησίας και έλεγχος ποιότητας εργαστηριακών μετρήσεων της μελέτης διαφορικής εξάτμισης ρευστών πετρελαϊκών ταμιευτήρων", Διπλωματική Εργασία, Σχολή Μηχανικών Ορυκτών Πόρων, Πολυτεχνείο Κρήτης, Χανιά, Ελλάς, 2021 https://doi.org/10.26233/heallink.tuc.88573
Appears in Collections

Summary

The objective of this Diploma Thesis is the calculation of the molecular weight of the residual oil (MW residual) from the PVT data of a Differential Liberation study, and the development of an appropriate correlation model, for predicting the value that the molecular weight of residual oil should have, for any oil based on the information contained in a PVT data base. It is claimed that the comparison of the two residual oil molecular weight values, is a reliable indicator of the reliability of the laboratory measurements performed during the study. The value of the molecular weight of the residual oil, is not included in data provide by the lab. However, it is a necessary parameter for checking whether the material balance that is expressed with available data is respected.Firstly, from the PVT reports of the 112 oils from all over the world contained in the data base, the molecular weight of the residual oil was calculated in two ways for each oil and the arithmetic average was taken as the correct value. Next, the difference between the molecular weights of the STO and of the residual oil (ΔΜW) was calculated.Finally, an Artificial Neural Network was built which receives as input reported lab measured data and was trained against the 112 data base oils to provide as output for any new oil PVT study the expected ΔΜW. The results showed that the Artificial Neural Networks has learned the ΔΜW with sufficient accuracy, with an average error of ± 0.03 g/mol. It is proposed that, for any oil PVT report, the Artificial Neural Networks is ran to predict the molecular weight of the residual oil and this prediction is compared to the value which is back calculated from the material balance equations of the study.

Available Files

Services

Statistics