Pb and LBE: a technological comparison Alessandro Gessi, Mariano Tarantino, Pietro Agostini ENEA Cr Brasimone 40032 Camugnano, BO, Italy Matgen IV School, Santa Teresa, 21/9/2011 Introduction • The goal of this work is to compare critically LBE (Lead-Bismuth Eutectic) and Pb, as coolants for GenIV fast reactors. • The choice of Heavy Liquid Metals for a nuclear fast reactors, comes from several known advantages, both technological and nuclear. • Hystorically, LBE was the first choice, due to its very low melting point (125°) compared with Pb (327°C). • However, several esperimental evidences, gained in recent years, suggest the need of a deep analysis and comparison between LBE and Pb as coolants, expecially as far as technological issues are concerned. • This work is a comparison of the two, starting from basic properties and going through non metallic elements behaviours, (i.e. Oxygen), corrosion, of structural materials and related technologies. Part 1: thermophysical properties (rif. OECD –NEA HLM Handbook, Chapter 2, A.Gessi, V. Sobolev) 1850 495 Surface tension, (mN m -1) . Pb 445 LBE Miller, 1954 Semenchenko, 1961 Pokrovsky, 1969 Kazakova, 1984 Novacovic, 2002 Pastor Torres, 2003 Kirillov, 2008 Plevachuk, 2008 Recommended 420 395 370 Bi 1800 Sound velocity (m s -1) 470 Miller, 1951 1750 1700 1650 Bi 345 1600 320 1550 300 500 700 900 1100 1300 300 Temperature, (K) Dynamic viscosity (10-3 Pa·s) 4.0 Miller, 1954 Kutateladze 1958 3.5 Bonilla, 1964 Pb Holman, 1968 Kaplun, 1979 2.5 Plevachuck, 2008 Kirillov, 2008 2.0 Recommended 1.5 Bi 1.0 0.5 300 500 700 500 700 Temperature (K) LBE 3.0 Pb Pb-Bi(e) 900 Temperature (K) 1100 Stremousov, 1975 Kazys, 2002 Habayashi, 2005 Pb recommended Bi recommended Vegard's law Pb-Bi(e) recommended 1300 900 1100 Part 1: thermophysical properties (rif. OECD –NEA HLM Handbook, Chapter 2, A.Gessi, V. Sobolev) 18 Bi Pb 14 Pb-Bi(e) 10 Brown, 1923 Lyon, 1954 Mikryukov, 1956 Powell, 1958 Kutateladze, 1958 Nikol'skii, 1959 Imbeni, 1999 Kirillov, 2008 Plevachuk, 2008 WFL law Enthalpy increment (J kg-1) Thermal conductivity (W m -1 K -1) 200 150 100 Recommended 6 300 400 500 600 700 800 900 1000 1100 Pb Bi LBE 1200 50 0 300 500 700 900 1100 1300 1500 1700 1900 2100 Temperature (K) Temperature (K) Bulk modulus Pb Bulk modulus Bi Compressibility LBE Compressibility Pb Compressibility Bi 35 150 0.10 0.08 25 0.06 15 0.04 Volumetric CTE (10-6 K-1) Bulk modulus LBE Compressibility (GPa-1) Bulk modulus (GPa) 45 LBE Bi Pb 140 130 120 110 100 5 300 500 700 900 0.02 1100 1300 1500 1700 1900 2100 Temperature (K) 300 500 700 900 Temperature (K) 1100 1300 1500 Part 1: thermophysical properties (rif. OECD –NEA HLM Handbook, Chapter 2, A.Gessi, V. Sobolev) Volume change at melting and solidification: A detailed knowledge of volume changes in metals and alloys at their melting points is of critical importance in the understanding of solidification processes. • Solid lead. Similar to the majority of metals with the FCC crystal structure, lead exhibits a volume increase upon melting. At normal conditions a volume increase of 3.81 % has been observed in pure lead [Iida, 1988]. • The situation is more complicated for LBE freezing and melting accompanied by rapid temperature change. In the handbook of Lyon [Lyon, 1954] a 1.43 vol. % contraction of LBE on freezing with a subsequent expansion of the solid of 0.77 vol.% at an arbitrary temperature of 65°C has been reported. P. Agostini et al. [P. Agostini, 2004] and Zucchini et al. [Zucchini, 2005] showed that the consequences of LBE volume expansion by recrystallization could lead to severe damages to pipeworks. The numerical and experimental studies described show that over-stressing due to LBE recrystallization and expansion in containment vessels such as in the MEGAPIE target must be considered during the design phase of the containment structures and can be managed by means of engineering rules. To avoid over-stressing of structures it is proposed to redouce: • the height of each solid LBE layer, • the presence of internal structures, • the LBE yield strength. Part 2: Oxygen The solubility and diffusivity of Oxygen in Molten Pb and LBE are very similar. The goal of controlling and monitoring Oxygen is a common need. Solubility and diffusivity of Oxygen in LBE and Pb, cfr. T. Gnanasekaran, Liquid Metals and Structural Chemistry Division Chemistry Group, IGCAR Part 2: Oxygen sensors Sensor output Basic components Solid electrolyte Yttria stabilized zirconia (YSZ) Tubes with 4.5–4.8 mole% Y2O3 "Thimble" with 3 mole% Y2O3 Reference electrode Metal/metal-oxide like Bi/Bi2O3 and In/In2O3 with Mo wire as electric lead Pt/air using steel wire with platinised tip as electric lead Second (working) electrode The liquid Pb alloy Auxiliary wire or the steel housing of the sensor serves as part of the electric lead Voltmeter reading, E Measure of the chemical potential of oxygen in the liquid metal May in general depend on the specific combination of the sensor with a highimpedance voltmeter Ideal sensor/voltmeter system Ideal zero-current potential: Calculated oxygen concentration, cO: C1 and C2 are constants specific for the reference electrode Oxygen sensors for LBE and Pb are based on the same principles: galvanic cells using YZR as solid electrolyte. Recent experiments have shown commonalities between LBE and Pb behaviours Part 2: Oxygen sensors Part 2: Oxygen sensors Configuration of the working electrode Metallic sheath (austenitic steel) with Pt mesh Electric contact by pressing the electrolyte against the Pt mesh The contact with the mesh is established at the highest testing temperature Disadvantages are the different thermal expansion of YSZ tube and steel sheath (rupture of the mesh during cooling) and oxidation of the steel sheath at high temperature Pt wire fixed with Pt paste Allows for producing different thermoelectric voltages using different materials (wires) for connecting the Pt wire at the closed end of electrolyte tube with the sensor housing Electric contact with the electrolyte may degrade during thermal cycling Comparatively small area of electric contact gives rise to high electrolyte resistance the Part 2: Oxygen sensors Part 2: Oxygen sensors Characteristics Electrolyte thimble Seal between electrolyte and housing immersed in the liquid metal Glass ceramic sealant developed for compatibility with YSZ and steel (thermal), and with liquid Pb alloys (chemical) Reference electrodes: Bi/Bi2O3 3-YSZ with optimized mechanical properties Prototype for oxygen measurement in a depth of ~5 m below the surface of a liquid-metal pool (based on R&D by IPPE) Part 2: Oxygen sensors Part 2: Oxygen sensors Part 2: Oxygen sensors Part 2: Oxygen sensors Sensor 1, 6 m Sensor 2, 2 m Thermocouples Sensor 3, 4 m Part 2: Oxygen sensors Part 2: Oxygen sensors Sensor design scaled-up from experience in smaller experimental facilities Output significantly decreases for immersion depth > 1 m Improvements of signal transmission required for oxygen measurements in pool- type reactors Output of reference sensor Immersion depth L, м 1 2 3 4 5 1 Output of the sensor under investigation as a function of the immersion depth Т, °С V, m/s Еref, mV aref Е6 а6 470 470 480 480 480 480 0,25 0,25 0,25 0,25 0,25 0,25 117 119 120 132 140 148 1 1 1 0,8 0,5 0,4 120 113 102 91 83 141 1 0,4 Design and Testing of Electrochemical Oxygen Sensors for Service in Liquid Lead Alloys Part 2: Oxygen sensors Two-shell electric of the reference electrode with guarding potential Part 3: solid slags and black dust The issue of solid impurities, “black dust” and macroscopic slags, has been one of the most important topics in the frame of HLM activities and experiments. In fact, during the operation (with LBE) of the CHEOPE III, LECOR and CIRCE facilities at ENEA several problems (filters and pipes occlusions, loops’ malfunctions, gas piping's blocks) have been encountered. Formed impurities have been sampled and analyzed: the presence of a relatively high amount of G and B phases together with the 40wt% ca. Of Massicot and Litharge (PbO) suggests a complex formation mechanism. Also, a sampling method problem exist: analytical methods can determine the composition of the samples, but not quantitatively determine a possible “formation rate”. The use of adsorption filters in the liquid phase gave good results. The filtered part appeared to be enriched in PbO, confirming the selectivity of the filters. A deeper sealing's control coupled with gas inlet filtration minimized the phenomena in LBE. NO solid impurity has been observed in flowing Pb (CHEOPEIII last campaign), even after 10.000 hours of operation, nor any operational problem. A fibreglass filter has been used also in Pb, where a small amount of PbO has been measured. Outgas systems appear clean. Part 3: solid slags and black dust “Black dust” SEM image, CHEOPE III outgas pipe Solid slags over CIRCE free level Examples of microscopic “black dust” and macroscopic slags (1m ca.) Part 3: solid slags and black dust Compound Concentration PbO 40 wt% ca. LBE (g b phases) 50 wt% ca. Fe, Al, Cr 10 wt % Ca. Table 1 Composition of a slag in the CHEOPE loop, LBE, 400°C, outgas filter. Compound Concentration PbO 60 wt% ca. LBE (g b phases) 30 wt% ca. Fe, Al, Cr 10 wt % Ca. Table 2. Composition of the filtered particles, fiberglass adsorption filter in the liquid phase, LBE, CHEOPE III Compound Concentration PbO 15 wt% ca. Pb 80 wt% ca. Fe, Al, Cr 5 wt % Ca. Table 3. Composition of few filtered particles, fiberglass adsorption filter, liquid phase CHEOPEIII, Pb, 500°C. Part 3: solid slags and black dust (* P. Turroni et Al., J.Vac. Sc. Tech. A 22(4)). Experiments performed in the frame of the TRASCO program: evaporation rates vs temperature. Part 3: solid slags and black dust The observed mechanism of solid impurities (gas and liquid phase) can be summarized as follows: Uncontrolled cold area on the facility Air pollution (ingas pollution) (2Pb+O2 samples. 2PbO) LBE recrystallization-phase separation (In the cold leg of LBE loops, T=350°C) Particle formation-macroscopic slags (reducing gas mixture bubbling is not effective) Loop draining-cooling down (samples are taken at room temperature in air) In the CHEOPE III loop Pb operated, where T=500°C and the maximum DT with the cold leg is 80°C, no slags or black dust has been observed. An indirect confirmation of this speculative mechanism is the recrystallized LBE found in the filters: it is not Pb+Bi but Gamma and Beta phases (Pb7Bi3 and Bi99,9Pb), suggesting a rapid cold point freezing. The formation of “black dust” happens ONLY with LBE. Part 4: corrosion The need for data on reference structural materials in contact with HLM is a crucial issue in the development of GenIV technologies. Lead and LBE are two highly corrosive media. The possibility to protect them by means of in situ passivation or artificila protections are widely studied in the frame of european programmes Corrosion mechanisms are driven by the same principles, both in LBE and in Pb. Elemental solubilities can generally be considered similar. However, given the higher temperatures of a Pb cooled reactor, corrosion phenomena are generally worse. Protecting steels from corrosion by means of in situ passivation is quite straighforward in LBE at 400°C, extremely tricky and less effective in pure Pb, at 500°C. in the latter, corrosion happens by means of mass transfer more than elemental straight dissolution. Part 4: corrosion T91 exposed to LBE, 3.000 hours of experiments, 500°C, Oxygen 10-6wt%. Thick protective oxide scales. Part 4: corrosion T91 exposed to Pb, 10.000 hours of experiments, 500°C, Oxygen 10-6wt%. Weak, thick, quickly formed oxide scales, easily eroded by HLM flux. FPN FIS ING Part 4: corrosion Fe: 89.5 wt% Cr: 8.3 wt% Fe: 71.4 wt% Cr: 8.4 wt% O: 18.5 wt% Fe: 41.5 wt% Cr: 12.5 wt% O: 42.9 wt% Fe: 57.0 wt% Cr: 0.4 wt% O: 41.3 wt% 20 mm scale micrography: oxide layers with corresponding EDS spots FPN FIS ING Part 4: corrosion 10.3 µm The coating scale have a very good continuity; Oxygen precipitation is observed below the FeAl coating; Small damages are observed in the coating maybe due to the post examination analysis; Part 4: corrosion 16 µm 33 µm Part 4: corrosion Old experiment at 400°C and latest experiment at 500°C. s Is the corrosion depth in microns FPN FIS ING Part 4: corrosion Corrosion curves for old and new experiments. Few points do not allow a critical comparison. FPN FIS ING Conclusion • The choice between LBE and Pb as coolants for GenIV fast reactor is connected to several open points: 1. Technological advantages and disadvantages (i.e. melting point, volume expansion, solid impuririties production, higher temperatures for structural materials) 2. Commercial issues, expecially Bi cost and natural abubdance 3. Nuclear safety issues, expecially Po210 aerosols production by irradiated Bi. The global amount of Polonium is produced only by Bi. With pure Pb, only the Bi traces are responsible of the eventual Polonium aerosol. The protection of structural materials from high temperature corrosion is thus the critical open point for Pb LFR technologies. Once solved, Pb could be the winning choice over LBE.