Assessment of structural damage due to cryogenic spill for FLNG plants: Beregning av konstruksjonsskade som følge av kryogenske utslipp på flytende LNG anlegg

Abstract

To investigate the effect of cooling by liquified natural gas, literature on the topic wasinvestigated in prior project work. Test results for cooling of steel by liquid nitrogen wascompared to numerical solutions of heat transfer equations. A simple 1D approach onheat gradients through a steel plate was developed for theoretical purposes, and over timeshow a comparable temperature profile with the test results. Test results comprised thecooling of a steel plate, from a trough placed over parts of the plate.Structural software USFOS, through the utility FAHTS, was used to model spill overa finite element (FE) model. Two models were utilized, a simplified deck plating modelgenerated specifically for this thesis, and a topside framework module - educational modelfrom NTNU. To describe the steel sections, we assessed steel in terms of performance whensubjected to cooling. Essential input to the analysis, in terms of properties for cryogenicfluid and steel, was identified prior to implementation in FAHTS and USFOS.Formulating the scope, the intent was to assess cryogenic spill simulations performed byComputIT, through the fire simulations software KFX. This being a new field of study,extending cryogenic spill with KFX as an extension to FAHTS proved to be too timeconsuming for the third party software provider. We chose to simplify our spill simulation,through FEM spill formulation for FAHTS, provided by USFOS AS. Applied spill routinestrictly assigns temperature to FE members, and does not offer a dynamic spill scenario.The spill routine, at beta stage for FAHTS, was applied to the best possible extent withinits limitations. Spill over different types of theoretical FEM elements had various compatibilitywith the spill routine, but the structural response in USFOS could be computed interms of desired temperature profiles from FAHTS. Restrictions were discovered duringthe analysis, and all results and interpretations adopted to these restrictions. Choice ofspill scenario was made from simple deductions on placement and function relative to anFLNG, and can be seen as specific to this thesis.Concluding on our results was done according to scope, with emphasis on steel performance.Important factors for the case of cryogenic temperatures was weighed in wheninterpreting the structural response.Embrittlement is the single most important factor in our analysis. Marine grade steelwill experience a ductile-to-brittle transition prior to cryogenic temperatures; performingslightly better than ordinary carbon steel. Disregarding any other factors, at the onset ofyielding, there will be abrupt fracture. On the contrary, cooled steel will have an increasein yield stress and tensile strength.Fracture will not occur from stresses imposed by cooling, but in connection with embrittlementand fatigue. Cracks at a critical crack length will not see any additional crackresistance, and there will be no crack arrest following added loading. Effects from contractingsteel, and the stress field arising to compensate for this behavior should be seenas the second most important consequence of cryogenic spillage. Being geometry specific,this does not exhibit any significant stress fields in this thesis.