Chemistry control of heavy liquid metal coolants in GenIV type reactors (PhD)

For some GenIV reactor systems heavy liquid metal coolants such as liquid lead or liquid lead-bismuth eutectic (LBE) have been proposed. One of the issues that arises with the application of these coolants is the fact that they are corrosive for the structural materials such as AINSI 316 or T91 steels that are proposed to be used. The underlying reason is that nickel and iron (albeit to a lesser extend) have a high solubility in the liquid metal at the operation temperatures (400-450°C) of the system. On the surface of the structural materials a protective oxide layer, that consist of an inner so called spinel layer which is essentially an Fe-Cr-0 compound and the outer magnetite layer is formed. In order to prevent degradation of the submerged components it is important that the protective oxide layer on the surface is not destroyed. The latter will happen when the oxygen concentration in the melt is too low. On the other hand the oxide concentration in the liquid metal should not be too high either. In this case, oxidation of the materials occurs resulting in a fast growth of the magnetite oxide layer. A too thick layer has negative effect on the thermal conductivity of the components which is particularly important for the fuel cladding and the tubing in the heat-exchangers. In additional item with thick magnetite layers is that flaking can occur. The poorly soluble magnetite that is then released in the coolant  can accumulate at cold spots (e.g. in the HEX and potentially cause a blockage. A second effect of high oxygen concentrations is that precipitation of PbO starts which can also give rise to possible blocking. Consequently, it is clear that for the long term reliability of these systems a control of the oxygen levels in the coolant is needed.

In order to determine the oxygen content of the liquid metal oxygen sensors exist. These are based on the chemical potential that builds up over a ceramic separator between the liquid metal and a metal-metaloxide reference. Industrial applications of oxygen sensors are obviously found in metallurgy but also e.g. in the production of flat glass where the liquid glass is poured on top of a liquid tin bath. The oxygen content of the liquid metal bath must be controlled to prevent formation tin oxide on the surface since this would be visible on the glass

The oxygen sensors that are envisaged to be used in liquid metal cooled reactor systems are similar to those employed in industry. Nevertheless some clear differences exist that require further development. Firstly, it is important that the oxygen sensors can also function in the low temperature region of the metal cooled reactor. Current oxygen sensors work properly at temperatures above about 400°C. The first objective of the PhD work is to extend the operation range temperature of the oxygen probes to lower temperatures where the target temperature is set to 300°C. This must be done by improving the signal strength and the response time of the sensor at lower temperatures. One way of doing so is to increase the transport of charge carriers through the ceramic. The second major difference between current industrial sensors and those envisaged to be used in liquid metal cooled reactor systems is the issue of reliability. It is clear that inside a nuclear system, sensors must deliver a reliable signal and should have an extended life span since replacement can only be done during maintenance periods. The main limiting factor in the lifetime of the oxygen sensors is failure of the ceramic separator which is mostly due to thermal of mechanical stress. The second objective is to improve the reliability of the oxygen sensors.

As described above, an oxygen sensor essentially consist of a ceramic separator, metal-metaloxide reference and an electrode that is used to measure the potential that builds up over the ceramic. In the optimization process these three components can essentially be optimized independently, provided that compatibility requirements are taken into account. The most important factor however is the composition and the shape of the ceramic separator. In the present sensors yttrium stabilized ziconia (YSZ) are mostly used although other possibilities such as stabilized calcia must be studied. 

The second relevant parameter is the metal-metaloxide combination. In the past both liquid metal (Bi, In, Sn) and fine powder metals (Ni) have been used. The former have the advantage that due to their liquid state at the operation temperature of the sensor, they form a better contact with the ceramic. On the other hand, frequent melting/solidification cycles seem to have a negative effect on the lifetime of the probe. 

As a final, issue, it must also be investigated if, and to what extend the requirement of operation at lower temperatures and thus a better permeability of charge carriers an the one hand and  better shock resistance on the other hand are competing  demands.

In order to realize these goals a collaboration was set-up with a small scale oxygen sensor producing company (Redox) that has the capability to produce ceramic sensor components on demand with given specifications thus creating room for research and development.