Multiscale thermo-hydraulic modeling of cryogenic heat exchangers

Abstract

The cryogenic industry has experienced a continuous growth in the last decades, partially sustained by the worldwide development of Liquefaction of Natural Gas (LNG) projects. LNG technology provides an economically feasible way of transporting natural gas over long distances, and currently accounts for nearly 30% of the international trade of this resource. The economic feasibility of these projects, in terms of both capital and operating costs, is to a large extent controlled by the performance of the main cryogenic two-phase flow heat exchanger. This industrial scenario provides then the motivation for a detailed study of the heat exchanger from a design perspective. On the one hand, it is widely accepted that a highly detailed analysis is required at a micro scale to properly take account of the two phase heat transfer process. On the other hand, a process-level description corresponds to larger time and space scales. In general, determining the proper methodology for considering these scales and their interaction remains a challenging problem. For this reason, current techniques focus in only one particular scale. The main objective of this project is then to develop a multiscale model applicable for two-phase flow heat exchangers. In this context, a three-scale framework is postulated. This thesis was divided into macro, meso (medium) and micro scale analysis. First, a macroscopic analysis provides a broad description in terms of overall heat transfer and pressure drop, using simple models without taking into account the details of physical phenomena at lower scales. Second, at mesoscale level, flow in parallel channels is considered following a homogenization approach, thus including the effects of flow maldistribution and partial mixing. Third, the microscopic description conceives a phenomenological representation of boiling flows, following multifluid formulations, for two specific flow patterns: annular-mist and post-dryout regimes. Finally, a multiscale design algorithm is proposed.