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

Search

Browse

My Space

Performance comparison between various commercial simulators

Mirkovic Aleksandar

Full record


URI: http://purl.tuc.gr/dl/dias/D2DEEE66-8114-489F-AC21-04181237A41C
Year 2018
Type of Item Master Thesis
License
Details
Bibliographic Citation Aleksandar Mirkovic, "Performance comparison between various commercial simulators", Master Thesis, School of Mineral Resources Engineering, Technical University of Crete, Chania, Greece, 2018 https://doi.org/10.26233/heallink.tuc.79834
Appears in Collections

Summary

This diploma thesis project will focus on the comparison of two black-oil simulation software. The simulators under study are the widely known Eclipse 100 and the relatively new tNavigator ones.All commercial simulators available in the market are supposed to solve the same differential equation problem using the same approach (finite volumes). However, the performance of them might exhibit significant differences even treating exactly the same simulation problem. Those differences might concern the computing time and/or the obtained results.In this work we started with simple Cartesian reservoir models and compared their performance against the simulators under study. Since the starting models were of low complexity and the background formulation of both simulators was supposed to be the same the results in terms of run time and simulators predictions should match. By verifying that it was shown that conditions for fair comparison were established and the next comparison step was performed with models of higher complexity. This allowed those fine differences in simulators to impact the prediction outcome, as well as runtime of simulation. As the general comparison strategy of increasing the model complexity is suitable for executing convergence study, this was done along the way. So, became able to evaluate how these two simulators are dealing with error introduced by spatial and temporal discretization as a part of numerical approximation of flow governing equations.The output data consisting of key parameters (oil/water production rate, average reservoir pressure and cumulative productions) for the respective model were compared along with CPU runtimes.By finalizing this study, a more solid judgement on performance of these two simulators on the field of black-oil simulations has been built.

Available Files

Services

Statistics