Vis enkel innførsel

dc.contributor.advisorLøvås, Terese
dc.contributor.advisorLien, Morten Stenstad
dc.contributor.authorBjørnæs, Marthe Hove
dc.date.accessioned2016-08-26T14:01:32Z
dc.date.available2016-08-26T14:01:32Z
dc.date.created2016-06-03
dc.date.issued2016
dc.identifierntnudaim:14961
dc.identifier.urihttp://hdl.handle.net/11250/2402296
dc.description.abstractThis thesis investigates how to optimize the calculation of thermal radiation from an advanced flare system in the numerical computation tool Kameleon FireEx (KFX). Measured radiation levels from a real flaring scenario at an offshore installation are used as reference. The flare system consists of one high pressure flare and one low pressure flare. An optimal approach for calculating radiation levels implies a trade-off between accuracy of the solution and numerical simulation time. To obtain a high degree of accuracy the size of the release cells and the nearby cells must be sufficiently small to capture the large gradients close to the release. This makes requirements of how the release cells should be defined and different configurations have been tried. The first cases were simulated using only one release cell for each flare which gives a low resolution. The resolution has then been refined by dividing the release cells into multiple smaller cells. There is a clear connection between the refinement of the release cells and the accuracy of the calculated radiation levels. The smaller cells are used, the greater agreement there is between simulated and measured radiation levels. Smaller cells lead to longer simulation time but this can be reduced by removing cells other places in the calculation domain. As the process equipment usually existing on the platform deck is removed in this specific scenario as a simplification, this allows for larger cell sizes far away from the flares. The conclusion of this study is that using 9 release cells for the high pressure flare together with 4 release cells for the low pressure flare was the most optimal approach for the scenario simulated in this thesis. It was in this case very low deviation between simulated and measured radiation levels, and the simulation time was considered acceptable. This configuration might not be the most optimal for any given flare; therefore sensitivity tests regarding release cell size should be conducted for other scenarios. The general conclusion is nevertheless that a high resolution of the release cells will give more accurate radiation levels.
dc.languageeng
dc.publisherNTNU
dc.subjectProduktutvikling og produksjon, Termisk energi
dc.titleNumerical Study of Radiation Hazard from Advanced Flare Systems at Process Plants
dc.typeMaster thesis
dc.source.pagenumber84


Tilhørende fil(er)

Thumbnail
Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel