Developing empirical heat-transfer correlations for supercritical CO2 flowing in vertical bare tubes

Abstract

This paper presents an analysis of three new heat-transfer correlations developed for SuperCritical (SC) carbon dioxide (CO2) flowing in vertical bare tubes. These correlations were developed from the large set of experimental data obtained at Chalk River Laboratories (CRL), AECL (2003). The dataset consists of tests performed in upward flow of CO2 inside 8-mm ID vertical Inconel-600 tube with a 2.208-m heated length. Data points compiled lie within a wide range of thermodynamic parameters – outlet pressures ranging from 7.4 to 8.8MPa, mass fluxes from 900 to 3000kg/m2s, inlet fluid temperatures from 20 to 40°C, and heat fluxes from 15 to 615kW/m2; and for several combinations of wall and bulk-fluid temperatures that were below, at, or above the pseudocritical temperature.The objective of our research is to obtain reference dataset on heat transfer in SCCO2 and improve fundamental knowledge of the heat-transfer processes and handling of supercritical fluids. In general, heat-transfer process to a supercritical fluid is difficult to model, especially, when a fluid passes through the pseudocritical region, as there are very rapid variations in thermophysical properties of the fluid. Thus, it is important to investigate SC-fluid behaviour within these conditions.In general, SCCO2 was and is used as a modelling fluid instead of SC water due to its lower critical parameters. Also, SCCO2 is proposed to be used as a working fluid in the Brayton gas-turbine cycle as a secondary power cycle for some of the Generation-IV nuclear-reactor concepts such as Sodium-cooled Fast Reactor (SFR), Lead-cooled Fast Reactor (LFR) and Molten-Salt-cooled Reactor (MSR). In addition, SCCO2 was proposed to be used in advanced air-conditioning and geothermal systems.Previous studies have shown that existing correlations deviate significantly from experimental Heat Transfer Coefficient (HTC) values, especially, within the pseudocritical range (for both SCCO2 and SC water data–see Section 4 for more details). Moreover, the majority of correlations were mainly developed for supercritical water, and our latest results indicate that they cannot be directly applied to SCCO2 with the same accuracy.There is still a great reliance on 1-D correlations for basic heat transfer calculations. We need to develop better empirical models that can predict SCF thermodynamic/heat-transfer behavior more accurately. Therefore, new empirical correlations to predict HTC values were developed based on CRL – SCCO2 dataset. These correlations calculate HTC values with an accuracy of ±30% (wall temperatures with accuracy of ±20%) for the analyzed dataset.