Library Subscription: Guest

ISSN Online: 2377-424X

ISBN Print: 978-1-56700-474-8

ISBN Online: 978-1-56700-473-1

International Heat Transfer Conference 16
August, 10-15, 2018, Beijing, China

IDENTIFICATION OF THERMAL CONDUCTIVITY MIXING MODELS FOR HETEROGENEOUS FOULING DEPOSITS

Get access (open in a dialog) DOI: 10.1615/IHTC16.ctm.023965
pages 4057-4064

Abstract

Fouling deposition on heat transfer surfaces is a costly problem in pre-heat trains of crude distillation units. Organic and inorganic species found in the deposits are characterized by markedly different thermal conductivity. Fouling heat transfer resistance depends on the extent of inorganic and organic phases, their structural arrangement, and stratification or layering. Understanding and, if possible, detecting the deposit composition would provide additional capabilities in monitoring, troubleshooting, prediction and optimisation of mitigating remedies. This paper aims at elucidating possible arrangements of the inorganic and organic phases by testing various thermal-conductivity mixing models for heterogeneous materials. A novel model describing the deposit layer as a multi-component structure is used with available plant measurements (pressure drop, temperatures, flowrates) to infer the deposit properties (e.g. thickness, conductivity and composition) over time. Various "mixing" models that describe the arrangement of the organic and inorganic phases in fouling deposits are tested. Radial composition profile of the deposit calculated with various mixing models is compared to the experimental composition analysis of deposits collected upon dismantling of the exchanger after four years of operation. The differences between this approach and a traditional analysis based on fouling resistance (that neglects deposit stratification) are discussed. Results show that the four mixing models tested provide an appreciably different level of agreement between compositions calculated from the plant measurements and the experimental characterization. The best agreement with the experimental data was obtained with a Co-Continuous model (all phases continuous), followed closely by an Effective Media Theory model (all phases discrete).