skip to main content

Title: Liquid Salts as Media for Process Heat Transfer from VHTR's: Forced Convective Channel Flow Thermal Hydraulics, Materials, and Coating

The goal of this NERI project was to perform research on high temperature fluoride and chloride molten salts towards the long-term goal of using these salts for transferring process heat from high temperature nuclear reactor to operation of hydrogen production and chemical plants. Specifically, the research focuses on corrosion of materials in molten salts, which continues to be one of the most significant challenges in molten salts systems. Based on the earlier work performed at ORNL on salt properties for heat transfer applications, a eutectic fluoride salt FLiNaK (46.5% LiF-11.5%NaF-42.0%KF, mol.%) and a eutectic chloride salt (32%MgCl2-68%KCl, mole %) were selected for this study. Several high temperature candidate Fe-Ni-Cr and Ni-Cr alloys: Hastelloy-N, Hastelloy-X, Haynes-230, Inconel-617, and Incoloy-800H, were exposed to molten FLiNaK with the goal of understanding corrosion mechanisms and ranking these alloys for their suitability for molten fluoride salt heat exchanger and thermal storage applications. The tests were performed at 850C for 500 h in sealed graphite crucibles under an argon cover gas. Corrosion was noted to occur predominantly from dealloying of Cr from the alloys, an effect that was particularly pronounced at the grain boundaries Alloy weight-loss due to molten fluoride salt exposure correlated with the initialmore » Cr-content of the alloys, and was consistent with the Cr-content measured in the salts after corrosion tests. The alloys weight-loss was also found to correlate to the concentration of carbon present for the nominally 20% Cr containing alloys, due to the formation of chromium carbide phases at the grain boundaries. Experiments involving molten salt exposures of Incoloy-800H in Incoloy-800H crucibles under an argon cover gas showed a significantly lower corrosion for this alloy than when tested in a graphite crucible. Graphite significantly accelerated alloy corrosion due to the reduction of Cr from solution by graphite and formation on Cr-carbide on the graphite surface. Ni-electroplating dramatically reduced corrosion of alloys, although some diffusion of Fe and Cr were observed occur through the Ni plating. A pyrolytic carbon and SiC (PyC/SiC) CVD coating was also investigated and found to be effective in mitigating corrosion. The KCl-MgCl2 molten salt was less corrosive than FLiNaK fluoride salts for corrosion tests performed at 850oC. Cr dissolution in the molten chloride salt was still observed and consequently Ni-201 and Hastelloy N exhibited the least depth of attack. Grain-boundary engineering (GBE) of Incoloy 800H improved the corrosion resistance (as measured by weight loss and maximum depth of attack) by nearly 50% as compared to the as-received Incoloy 800H sample. Because Cr dissolution is an important mechanism of corrosion, molten salt electrochemistry experiments were initiated. These experiments were performed using anodic stripping voltammetry (ASV). Using this technique, the reduction potential of Cr was determined against a Pt quasi-reference electrode as well as against a Ni(II)-Ni reference electrode in molten FLiNaK at 650 oC. The integrated current increased linearly with Cr-content in the salt, providing for a direct assessment of the Cr concentration in a given salt of unknown Cr concentration. To study heat transfer mechanisms in these molten salts over the forced and mixed convection regimes, a forced convective loop was constructed to measure heat transfer coefficients, friction factors and corrosion rates in different diameter tubes in a vertical up flow configuration in the laminar flow regime. Equipment and instrumentation for the forced convective loop was designed, constructed, and tested. These include a high temperature centrifugal pump, mass flow meter, and differential pressure sensing capabilities to an uncertainty of < 2 Pa. The heat transfer coefficient for the KCl-MgCl2 salt was measured in two different diameter channels (0.083 and 0.370Ã). In the 0.083 channel, the experimental heat transfer coefficient was shown to agree with values obtained from heat transfer correlations used for water. In the 0.370 D channel, the experimental heat transfer coefficient data was predictable by either a correlation for mixed convection, or forced convection depending on the value of Gr*/Re. These experiments provided new insights into the construction and operation of molten salt flow systems. The selection of multi-component salts for molten salt flow systems requires knowledge of properties such as melting point, heat capacity, density, and viscosity of these salts. Theoretical models have been developed for the prediction of these properties of multi-component salts.« less
; ; ;
Publication Date:
OSTI Identifier:
Report Number(s):
DOE Contract Number:
Resource Type:
Technical Report
Research Org:
University of Wisconsin, Madison
Sponsoring Org:
DoE - Idaho Operations
Country of Publication:
United States