 Aujourd'hui, nous continuons de faire la présentation de Dr Cristian Lajer. Il a fait plusieurs changements, des petits changements. Cristian, s'il vous plaît. Bonne matin à tous et à toutes. Il y a un changement que j'explique à vous. J'ai adressé la première partie de ma présentation d'aujourd'hui, mais j'ai découvert, je dirais, la plupart des pollutions. Donc ça semble être plus logique maintenant de parler du contrôle de la qualité sodium. Ok, pendant une heure et demi. Et puis je vais revenir à la deuxième partie de ma présentation d'aujourd'hui, à la fin de la matin. Et des matériaux ce matin. Si vous agreez. C'est juste d'avoir plus de progrès logique. Je suis désolé pour ça. Donc, ok. Et si vous ne vous inquiétez pas, parfois je vous le verrai devant mes screens. Merci. Donc, comme vous le savez, pour chaque coulotte, en fait, c'est assez nécessaire pour adresser le contrôle de la qualité de la coulotte et de l'implification, pour demander des requiredments pour satisfaire des requisites et alors, c'est très important pour chaque coulotte. Alors ici, je vais adresser le cas certain sur le Sauvignon Fast réacteur. Mais si vous avez des questions sur le lead, par exemple, on peut aussi avoir un court discussion sur ce point. Pour l'esphère, c'est généralement pour l'esphère principale, j'ai écrit Sx, s'est-ce que l'esphère peut être S pour sodium ou L pour le réacteur rapide, ou pour l'esphère de l'esphère de l'esphère. L'esphère est un paramètre qui est un paramètre de la corrosion. En termes de conséquences, il y a deux kinds de conséquences d'oxygène, il peut être une corrosion, donc ça signifie la réduction de la séquence de l'esphère, ou en général, évidemment, la contamination, parce que nous produisons des produits de corrosion activés, et nous pouvons induire des dosimétries. Donc, la dosimétrie n'est pas peut-être une issue, une grande issue, si vous n'avez pas de maintenance operation, mais c'est clair que si vous devez, let's say, remove intermediate heat exchanger, or any component for repair, for example, you have to deal with this contamination, and this dosimetry, consequently, this dosimetry, so necessity to then contaminate, handling, repair. It can be also for in-service inspection, also sometimes you need to be able to be near, even if we have now some strategies to have access to the welds, for example, to be controlled without any hazards of any constraints due to the contamination. So, generous necessary to master the oxygen. Oxygen weld master can help to maintain oxide layer stable. This is for lead, particularly heavy liquid metals, because in order to limit liquid metal embritement, which is a phenomena which can be deleterious for the integrity of the components, it's necessary to maintain this oxide layer, even if there is some other options, some other options like coatings, for example, you know that for materials, for cladding, for every liquid metal, it's possible to aluminize, for example, the cladding with, let's say, geyser process developed in Karlsruhe, and of course, but oxide is a protection, and you know why, because in fact, when you have oxide, you have less wetting and so less interaction between the liquid metal and the structural material. Oxygen can induce also precipitation of some coolant oxide, for example, for heavy liquid metal, you can have precipitation of lead oxide, because it's due, of course, we have effect of solubility, but also the lead oxide has a low lead oxide in the lead or lead business coolant. You have very, I must know, dissolution, in fact, dissolution in sodium. Instead of, for sodium, for example, for sodium, due to the fact that the sodium is a reducing element, as soon as you have injection of sodium oxide, you have a dissolution in the sodium. And so it means that the oxygen in a sodium is not as a particulate, generally, it has dissolved, and so you have to reduce the concentration in oxygen, which is dissolved in the sodium. So it's quite different situation than with heavy liquid metals. And also in some cases, you can, if you have oxygen, if you have binauraloys like lead bismuth, or in another case, not for here, but lead lithium, for example, if you have oxidation, one of the two compounds, okay, like lead or lithium in the case of lead lithium, you can modify the composition. So it means that you need to introduce maybe the element, the compound, which is oxidized. So there are some consequences of this oxygen. For intermediate circuits, and this is typically only the case for sodium fast reactor, I recall you that we foresee in the sodium fast reactor an intermediate circuit, just only because we want to avoid any hazards of interaction between the primary sodium, which is activated, okay, with sodium 22 and sodium 24, you remember yesterday, my lecture. And so for this reason, we need to avoid a potential interaction between this primary sodium and the water with production of gas, and as you know, in a reactor, we absolutely avoid the transfer of gas bubbles in the core in order to avoid any reactivity issues in the core, okay. So this is a point. But for intermediate circuits, we have to deal with, I will explain later, hydrogen produced by the coming from the continuous corrosion in the steam generator. We have in any steam generator, it's not specific of sodium. It's for all systems equipped with a ranking cycle, energy conversion system based on a ranking cycle, and you produce by corrosion on the water side, some hydrogen, magnetite and hydrogen. And for this reason, we have an introduction of hydrogen in the intermediate circuit. But we will see later that thanks to this hydrogen, and by trapping this hydrogen in the secondary loop circuit, we trap also the tritium produced in the primary by a ternary fission of plutonium or coming from the boron carbide. So due to that, it's necessary if you want to detect very quickly a sodium-water interaction due to ingress of steam in the intermediate circuit, it's absolutely necessary to have in steady state operation a very low hydrogen content in order to detect very early the sodium-water interaction in the intermediate circuit. This is for all the circuits, of course. If you have a too high concentration, if you have a cold point, you can have crystallization of sodium oxide or end sodium hydride also in some places. And so if you want to avoid any plugging or any plugging of narrow gaps, for example, it's necessary to maintain to keep the oxygen and hydrogen content enough low in order to avoid any hazards related to this situation. OK. So now coming back to the sodium, we are going to check again, maybe I have already said some points on the way it's necessary to purify the sodium. I'm going to explain to you what we did on the basic phenomena of crystallization. For students, it's interesting because in the past, you will see that it was this component, what we call co-trap, was based only on a thermo-hydraulic and thermal approach. And in fact, it's necessary to have a more, I would say, chemical engineering approach in order to have a better design of this component. And I will show you that from this basic data and also from the operational feedback coming from the reactors like Phoenix or Super Phoenix and others, how we have developed new concepts taking into account the feedback and also the basic data like kinetics, in fact, kinetics of phenomena. And at the end, I will address some more specific points like how we do mass balance, mass balance of the traps, how to follow this component. And I will explain to you also what is the evolution of the design, what we did in the 80s, okay, you were maybe not born at this time, but how we proceed now taking into account all this data we have gathered. Okay, so just to recall, and I go in the first part first. Here you can see a scheme I have shown to you yesterday. You can see here introduction of, I would say, for example, fresh fuel with the cloud and the metallic surface, and you have an introduction of pollution, okay, because sodium is a reducing element. So, okay, this is exactly, here you can see what I explained to you. It's some kind of drawing showing on the right side in green pipe in the steam generator. Inside the pipe, you have the water, and you have a corrosion producing manetite and hydrogen. In case of leak, you have injection of water, production of hydrogen in the secondary circuit, but also in steady state operation inside the water. As I told you, corrosion, even if we try to limit this corrosion by using some products like hydrazine, for example, hydrazine is N2H4, okay, which is used to limit the oxygen level on the water side. Just recall that we have here the solubilities, okay. We use for the sodium, as I told you yesterday, for the oxygen, we can use a northern low. There are some different lows, which are very close, generally speaking. And on the secondary loop, we have a solubility low produced by Dr Whittingam in the UK. And it's generally a good reference. For oxygen, there are many solubilities lows for hydrogen, much less. So we have on the primary loop requirement to keep the oxygen below 3 ppm. 3 ppm, if you look here, corresponds to about 150 degrees Celsius in terms of saturation temperature. I will explain in more detail what is the saturation and solubility. And for hydrogen, generally, we accept below 0.1 ppm, 100 ppb of hydrogen. It seems very low, but in fact, it's rather easy to obtain, okay. There is no difficulty, particular difficulty. Of course, if you are in steady state operation, in case of sodium-water interaction, for example, it's clear that you can have higher values, okay. So, we have different pollution sources. One is some of them are continuous. So the diffusion of hydrogen through steam generator units works. And impurities in cover gas. Also, you can have a very, very small pollution. And as this continues, oxide dissolution in sodium, when you introduce a metallic surface. And in this case, it is clearly discontinuous because it's after when you introduce a certain number of fuel assemblies in the reactor. Of course, you have a pollution. Then, we will purify. And then, in steady state operation, you have no pollution, no additional pollution. Air ingress, after repair. In case, when you repair, sometimes you have not a perfect tightness of the system. And you need to take into account air ingress. Air ingress. And with air ingress, you have oxygen and moisture. So, you have some pollution. And after that, when you fill again the circuit with sodium, you have a pollution. Water ingress in a steam generator unit. Of course, when you have a sodium water interaction. Sometimes, we felt in the past oil ingress. It was the case, for example, in PFR. In PFR reactor in UK, in Scotland, exactly. There was some pollution with oil. When you have oil, interaction between oil and sodium produces tarves, hydrogen, methane, and so on. Of course, limited amounts, but you have some pollution of the sodium and also of the carbon gas with carbon products. Metallic filling due to maintenance. Yes, it happens. Sometimes, it was the case in Cayenne Cattu reactor in Germany. In Germany, there was an experimental reactor, which was closed at the beginning of the 90s. And there was some metallic filling due to maintenance. Main process for sodium purification now. When you have these impurities, of course, the two most important impurities are oxygen from oxygen or hydrogen. Hydrogen in the secondary circuit mostly. So, it is necessary first, as I told you yesterday, to obtain nuclear grade sodium. So, this is the business of a company who supplies the sodium. Okay, clearly. Of course, before to start the reactor, you have initial cleaning of the loop, components and vessels. And particularly, we circulate hot nitrogen, let's say, around 200 degrees Celsius, with two goals. One is to heat up all the structures. It's not an easy task because you have a lot of structures. And also, you dry, I would say you dry more or less the structures in order to reduce the pollution. Okay, so this operation has been done for all the reactors. I think it was recently done or currently done in India for PFR reactor. You know that in India, they are not finishing the manufacturing. I have seen that yesterday. It's commissioning phase and maybe they should start to fill up the reactor with sodium. Of course, at the very beginning, we have also filtering because when you have a so large amount of sodium, even if the sodium produced supplied by the company is clean, you have to filter. For example, for Super Phoenix, we have four levels of filters before the sodium to enter in the reactor. Coal trapping, this is the main point. This is the main process for these large size systems. We will see that also there is another technology. I would not address too much, but hot trapping. We developed hot traps for some very specific applications. We show you after. And of course, the key point is when you operate, particularly during the maintenance and so on, to limit the ingress of pollution by appropriate operating rules. You have to avoid any... You have to work with a very clean approach. Crystallisation. So our system for cold trapping is based on the crystallization of impurities. You know that there are three different technologies to promote crystallization. Why is liquid evaporation like salt? You maybe know that to produce a very good salt. So it's not... It could be in principle possible. You can extract the sodium from the reactor, evaporate distillate, I would say. And then recover the pure sodium and so on. Now it's not forget it. Okay, too complex. Reduction of the solubility by adding some additive impurity. Sometimes it happens and when you want to produce a beautiful crystal, it can be used. But in a nuclear system, it's difficult to imagine it. For sodium, I don't know which product we could have. But you modify by adding some products, the solubility law of impurities, and you promote the crystallization. So it's used in laboratories in chemistry, but not here. And the last point, which is used extensively, is cooling the liquid below saturation temperature. Of course, the last option is the most attractive. It's easy to implement. We need to have a cooling. And of course, to foresee solid retention. So important point is to understand solubility. Because if you understand this phenomena after that, it's quite easy to understand how we operate. Here you have a so-called diagram of Oswald and Meier, where you have a concentration as a function of temperature. Just I take an example of the morning. Okay, coffee. You introduce a piece of sugar in your coffee. Okay, you increase the concentration, of course. And what you can do is add a second, if you like, sugar. Two pieces of sugar. And nothing happens. And what happens if you increase more? Sometimes you have no more dissolution. So your piece of sugar remains in the sodium, but not sticky. So it means no dissolution. So if you want to recover your sugar, you can filter your coffee. Okay, you get back a part of the sugar, because one part is dissolved. Now, okay, you are serious. You are serious, and you say, okay, just only two, even if, okay, it's not too much. But two pieces of sugar, and instead to drink your coffee, you go to discuss with colleagues, with friends, and so. And what happens? What happens here? Do you know what happens? Yes, you have no crystallization, in fact. Because, in fact, you have the solubility curve here. But, in fact, you need to have additional sur-saturation to create the good conditions to have a nucleation. If you reach what we call the sur-solubility curve, you go through what we call metastable region. You reach a point where we have nucleation. Okay, nucleation. And it means that you create new small, very small crystals, nuclei, we call that nuclei, and you produce these crystals. If you stop here, in fact, if you are at this level with this concentration and this temperature, nothing happens. Except, except, if, except, if you had a piece of sugar, a crystal of sugar, it will grow. You have no new crystals, but you have a growth of this crystal. And, okay, here, what happens? But, in fact, you have a crystallization of sugar at the bottom. And in this case, you don't filter because the sugar stick at the bottom. You need to have some hot water to clean up. So, in this case, you have had what we call heterogeneous nucleation. It means that you have crystals, but these crystals need to have a support. Okay? A structural support to support the crystals. Okay? If you have understood this example, culturep is the same above. It never happened to me. Okay. So, in terms of nucleation, we are generally his antalp diagram with the energy and the nucleus size. In fact, when you have structure, you can have what we call heterogeneous nucleation. And it needs a lower level of energy to have this nucleation. If you want to have nucleation in the bulk, in the liquid metal bulk, you need more energy. But in practice, generally for sodium, we never reach the right conditions to have homogeneous nucleation. So, we don't have crystals of sodium or hydride of sodium oxide inside the sodium. We generally have only nucleation on the structural material. So, it's the first point. So, it means that the culturep is not really a filter. You provide the support, it's important, because sometimes people say it's a filter. No, it's not a filter. You provide the structure and the surface to promote heterogeneous nucleation. So, it's different. Okay. Nucleation phenomena. Okay. So, again. So, about the growth. In fact, we assume that... Sorry. Yes. Let's go to the first, this one. Here you have a crystal. Okay. You can see. We have a boundary layer and you have here the sodium circulation. And we have in the sodium a concentration. We have a temperature. And so, you can have a super saturation on this level. And so, in this case... Sorry. Yeah. You can have... You have... We assume that we have some boundary layer. We have a diffusion of the impurities through this layer. And then, a second step is the integration of these atoms. In fact, these ions, because we are in a reducing field. And you have integration of this impurity of these atoms, oxygen or hydrogen on the crystal, in the crystal lattice. Okay. In the structure of the crystal. Here you can see... show your different layers and increase the number of layers when you increase the... when you have growth. In fact, the kinetics of the overall phenomena is the kinetics of the lowest step. It means that it's like when you are going to the stadium, okay, a football. Sometimes you have a control. They control your backs and so on. So generally, you have a lot of... Okay, your kinetic to reach your seat in the stadium is limited by the control of the backs. Okay. And here, it can be the first step. And the second step can be if you run very fastly between the control and your seat, you can reach the time. So, we will see that there are two steps. And if you have a very slim, you can go through the... very quickly through the control. This is typically the case of hydrogen. Okay. Hydrogen is a small atom. Okay. And go through the control and reach, and after that, you can reach quickly your seat. Oxygen, which is more... It needs more time. So, it's the reason why, for example, in these two steps here, we will see. We have demonstrated, I will show you after, that the oxygen, the limiting step of the hydrogen is the diffusion through this boundary layer. And in this case, when a phenomena is limited by the dissolution, the order of the kinetic is one. Okay. And if it has been demonstrated also by more phenomenological studies and theoretical studies models, that the integration and the listes, the order is two. Okay. It's a model called... model of Burton, Cabaret and Frank, which was improved, but the order was never changed. Okay. So, in the case of... So, here you have the kinetics. Okay. Diffusion through a boundary layer. You have... Here, the kinetic is generally a constant, which is the function of the temperature also. This is an Arrhenius law, a reference surface, and the difference between the concentration in the bulk and the concentration at the interface. We assume that, effectively, we have a concentration here. Sorry. Yeah. We assume that we have a concentration. This is a concentration in Sonia. We have an intermediate concentration at this level, which is an equilibrium value. And it's... But you can understand that it is not easy to measure this concentration close to a crystal. So, this kinetics, this is the first kinetics. The second one is the integration step in the crystal solution interface with an order, which can be... which is generally two. Okay. And due to the fact, it is difficult to know intermediate concentration. We use what we call an overall equation. Okay. We integrate the two. And if the gross is limited by the diffusion step, then the kinetic and the order is one. And if the gross is limited by the integration step in the crystal lattice, the order is two. Okay. Here you can see the theoretical law. Here you can see a crystal of oxide in liquid sodium. Okay. So, you can see these layers. Okay. It's not just only a theory. It's also the same for sodium chloride, natural sodium chloride. We can see these shapes also at the microscope. Two gross phenomena. If you have a very low supersaturation, it means if you are near the solubility, you have a pyramid... you have what we call a regular gross and you have a pyramid. And you have, in case of high supersaturation, the dendritic gross. Okay. Very beautiful things you can observe in a contrappe. Okay. So, what we did, we made a lot of experiments in the 80s. And we measured the kinetics of nucleation and gross, only for sodium oxide with circuit just only operated with oxygen just only for hydrogen in a circuit only polluted with hydrogen. Okay. So, it was a complex approach and about to two different steps with our facilities. And we established the kinetics and particularly the energy of activation of the Arrhenius law and also the order of the kinetics. You can see here for the nucleation. On observe that, sorry, we will see just after. Yes. Here you have the kinetics. Okay. And here you have the order. Just... So, what we can notice is that for nucleation kinetics we have a large impact. If you look here, the activation energy for hydride is very high compared to the sodium oxide. Okay. Minus 60. And so, it means that there is a large influence of temperature for the sodium hydride nucleation. And for the gross, you can see that the two, as I told you, the story of the stadium. Okay. You can see that the order for oxide we have found is one. It means that the phenomena of gross is limited by the diffusion of the oxygen between the sodium burq and the crystal lattice. And for the hydride, it's quite different. The limiting step is not the diffusion, it's the integration in the crystal lattice. So, it means very theoretical. But now, we are going to see how we have used these results. Okay. Before that, I just recall the principle of operation of a cold trap. So, here you have always the diagram concentration, temperature. In a cold trap, generally it's a vertical position. Okay. But here, for explanation, we prefer to put the cold trap in a relaxed position. Okay. So, you have a horizontal. Here, you can see inlet of sodium. And then, we circulate in what we call exchanger economiser. In fact, the sodium cooled coming from the cold trap and going out, purified, helps to cool down the sodium entering in the cold trap. And here, you have an external cooler. Here, you can see. This is here, in this case, for example, a coil in which we circulate oil. Okay. Oil. And in order to exchange between this external jacket and the internal, this coil is inside a static sodium potassium alloy. Why sodium potassium? Because, as you know, sodium potassium is liquid at ambient temperature. So, it means if you cool down the cold trap, you have always liquid and you have not stresses, mechanical stresses on the structure. And so, at the end, so you cool down. And here, you reach the coldest point. Okay. And you create in this area. Okay. Oops. Yes. When you reach this point, you have not crystallization of sugar, clearly, but you crystal of oxide and hydride. Okay. And so, you have crystallization in this lower part. Here, you have some kind of mesh, a metallic mesh, generally used. In fact, you create conditions to create a lot of positions for nucleation of crystals. Yeah. And so, in this area, you have a mesh with a certain density, density. In other, if you want, it depends if you want to have more sites or less and so on. There is some design routes here. And so, the sodium go through and enter in the inner pipe and circulate here in the heat exchanger in order to cool down the sodium entering in the cold trap. Okay. Here, you can see a wire covered by crystals. Okay. And okay. So, what is important to notice is that there are two main parameters, operating parameters. The flow rate of sodium, of course. And the temperature. And it is clear that the cold point temperature is the key point. And this, what you need to know to fix this temperature, you need to you need to know at which temperature you promote crystals. For this purpose, we use an apparatus called the plug indicator. It's a system which allows you to detect, you cool down your fluid and you identify at which temperature you promote the nucleation which show you in more details a little bit after. So, you need an information about the concentration and then you fix, of course, you need to estimate these temperatures here. What we call plugging temperature means temperature at which we create crystals. And you need, of course, to fix the cold trap temperature. Let's say we have used 20 degrees lower because it corresponds to some kind of optimal conditions. Ok? So, this is the way we process. And so, and ok. So, this is the principle. Ok? So, what are the criteria we move to the design? What are the criteria is, of course, we are, what we call instantaneous efficiency. What is, if you have a very good cold trap, well designed, when you enter at the concentration, the concentration at the outlet, what is the minimum concentration is here, in fact, exactly here. Ok? You cannot be in the under-saturated area but you reach this point and then you will go out. But the kinetics is not instantaneous phenomena, the crystallization. So, it means, in fact, in the system which depends on the circulation, the flows and so on. You have, of course, generally in the past design the cold traps they never reach an efficiency of one. What is efficiency? So, variation of concentration is concentration at outlet minus concentration at the outlet, and the maximum is concentration at inlet minus the concentration, the minimum concentration you can reach is at this level. It is explained here. Ok? This is exactly, so this value, this ratio is between 0 and 1. The purification rate is the efficiency multiplied by the concentration at inlet minus the concentration corresponding to the cold point temperature multiplied by the flow rate. The capacity is in fact the feeling rate of each area in the cold trap. We will see that it can be the mesh, the wires of steel they vote to crystallize for the crystallization and sometimes cooling surface particularly for hydrogen. Ok? The capacity, it's important because you, this apparatus you feel it and at the end you have an evolution of the flow dynamics. We show you during the modeling later and it's important to it's important to know that we sorry I lose my it's important to design appropriately the different zones of the system. And the compactness, what is the compactness is just here you can see that you can have a cold trap where you have a very small part of the used for the impurities et with a very large component maybe 10 meters and 2 meters diameter to trap 5 liters of impurities. Ok? It's not optimized. So there is clear parameter in a reactor when you design the systems you need to have the best volume in a given volume to optimize the feeling rate of this volume and particularly for the primary circuits because as the sodium is active we want to avoid we, there is generally for an external cold trap inertization of the building and so in this case the cost and you could have a big impact on the design of the plant Ok? So generally the purification rate has to be the highest why it's important when you have a big reactor when you want to clean up a sodium a primary sodium for example if you say to the operator like EDF or NERSA for Super Phoenix I need 2 months to purify if 2 months of operation lost ok because your concentration is too high of course he's not very happy because 1 million euro per day lost so it means that clearly there is an economic impact of the design and there is a clear objective to have a very high purification rate capacity the highest value is the best for the design for the design of the building capacity has to be the so it depends of the strategy if you have a small cold trap you remove periodically or you can have a trap you implement for the world life of the reactor but don't forget that we trap not only impurities like oxygen and hydrogen but also we trap it's not a problem of volume of impurities but of contamination we trap also activated corrosion products and so you can accumulate some high very high dosimetry so in this case when you want to replace or for the decommissioning we could have an important impact on the cost of the decommissioning of this system which is I would say the kidneys the kidneys of the reactor ok here you have the kinetics ok the solubilities ok you have seen plugging meter to measure the oxygen content here you have in this part you have the circulation of sodium from we have inlet of sodium circulating in the space here ok when you arrive here this is we have a cooler here and at this level we have what we call a pellet ok with 12 grooves and you when you have when you reach the solubility super solubility you create nucleation and due to the fact you have a very narrow narrow holes you have a plugging ok so if you follow the temperature by cooling the sodium and if you follow the flow rate when you have nucleation you have a partial plugging of these holes and you have a reduction of the flow rate ok this is just an illustration here you have in red the temperature ok it's automatic in fact so you have a reduction you have a high temperature you have a reduction when you reach for example a certain value it can be we have we reach 110 because it's a minimum temperature I would say if we are below 100 there is a risk to freeze the sodium ok so we stay above ok but thanks to the solubilities when you reach this value you have an information about the concentration it's an estimation but if you are used for that I don't go in detail but it's a compression of the content ok and after that you have a plug in here you can notice you start the plug in here and the reduction of the we have a slope here so it means that you have precipitation of impurities in the narrow holes and then when you reach a minimum value of the flow meter of the flow rate why? because if not there is a risk to have a total plug and in this case if it is hydride it's possible to have a thermal decomposition at 300 but if it is a plug of oxide the decomposition of sodium oxide is above 900 degrees Celsius so it means that you need to replace put the plug indicator at the garbage and start with a new one ok so it's clear that we have a plugging temperature here but sometimes there is a good estimation of the concentration by measuring what we called unplugging temperature because unplugging it means that in the groove in the narrow holes you have some kind of equilibrium equilibrium between the growth and the dissolution also so it means that you are very near the solubility solubility is associated to the saturation temperature and thanks to the solubility laws you can know more accurately the concentration of the of the impurities ok and if you are we did that for example in a specific case in superphénix when we have a large pollution when you are near the unplugging temperature you increase the temperature you observe a dissolution you reduce again the temperature you observe a crystallization and so on so you can manually establish more accurately the concentration ok this is the plugging curves during air ingress in superphénix oops sorry we you know in superphénix we have an ingress of air in the primary circuit a large amount of air but due to the fact we had a very large amount of sodium 3,000 3,500 tons of sodium about we reach pollution we reach a value of 15 ppm instead to have less than 3 ppm we reach 15 for some operators for an operator they consider it's not too far not too high but in our case effectively we consider that it was not acceptable so we stop stop the reactor purify it in less 4 months about ok to reach a good value but we lose 4 4 modes of production of a reactor of 1200 megawatt electrical megawatt you lose a lot of money of course so it was an event in June 1990 and I was particularly involved in this activity to recover the quality of the sodium in order to restart the reactor so here you can measure so you can see that this it's not so clean as I show you previously as a plug-in temperature but here you can notice that here you have a steady state flow rate and then suddenly you start to have this plug-in ok so the plug-in was estimated to 218 ok what is the purification procedure purification procedure is when you know the plug-in temperature for example you have an idea of the saturation temperature we estimate maybe 20 degrees between the solubility and solubility curve and so when you know this first value generally you fix the cold point temperature here generally 20 degrees below so you purify of course if your cold trap works it's of course the case in the main circuit your concentration decreases here you can see decreases and when you have purification temperature you reduce again the cold point temperature and so on so you reduce progressively in order to reduce the oxygen impurities content in the sodium so it's a step by step procedure and when you stop you stop when the cold point temperature is around 110 ok not below because you want to avoid any hazards of freezing the sodium in this design some design options now sodium distribution is a cold trap you have a distribution it's what we call ring type cooling fluid can be hair, oil, sodium potassium and oil in static sodium potassium and so on internally take changer to avoid the plug-in in the inlet pipe as I explained you the cooler it can be a single cooler or modular coolers we have developed concepts with the modular coolers in order to have some flexibility to variation of the delta T in the cold trap as a function of the situations support for impurities it can be a knitted mesh as I told you some wires these wires we have a supplier providing wires the diameter of the wires 0.3mm so you can see it's a small diameter poly rings in Germany for example and we have developed also concepts with no mesh but just only a cooled wall why because near the wall where you have the cooler you have a delta T and as I told you in the kinetics it's very efficient to trap the hydride sodium hydride yes really location of support ok the location and the surface pavilion depends on the design requirements ok there are some requirements I don't go of course in detail here an example of cold trap which was very innovative in the 70s end of the 70s it's integrated cold trap this concept has been also recently selected for BN 1200 reactor in Russia and also for the next step for the FBR FBR reactors that are designed in India after the PFBR they intend to build the two FBR in India in Kalpaka and the same for here and also for VTR project even if now it's more or less frozen this project but they have foreseen and integrated you can see here the reactor ok here you have the primary circuit the core here the structure below the core sodium circulates here enters at about 400 degrees et leaves at 550 this is an example of super phoenix here you have a control plug where you have here the temperatures below this control plug in front of each outlet of fuel assemblies to detect any variation of temperature from the core any plugging in fact because if you have somewhere a partial plugging of course you lose your cooling efficiency so you have an increase of the temperature initiateur of civil accidents of course you can reach maybe boiling ok in one sub assembly and in this case you can create a sequence of civil accidents but here you have hot plenum here and the cold plenum below ok here the sodium enters hot sodium enters in the intermediate exchanger and then the sodium comes back in the lower part and the sodium is sent thanks to a mechanical pump below the core through what we call a diagrid a diagrid is a system which allows you to fix all the hexagonal fuel assemblies and constitute the core and for the first time we decided to have internal coal traps what is the coal trap here it is located in the hot plenum it means at the temperature of 550 at the bottom here you have a electromagnetic pump ok which extracts the sodium hot sodium from the hot plenum circulates here you can see in orange you have an internal heat exchanger then the sodium reaches a temperature of 180 and then is sent inside here you have here you can see pink coil it is in which we circulate nitrogen of course we are not going to have air in the primary circuit ok so nitrogen and so we have cooling the cooler is here and when the sodium is extracted a part can be directly released or another part can circulate in the heat exchanger why a part only is if you have a different flow rates in the heat exchanger you have the capacity to modify the temperature the entrance of the cooled area and what is interesting here you can see here on the upper part above the slab ok you have here an operator you have a hole in which we can introduce what we call a cartridge the cartridge is inside so in this cartridge you have a mesh to support the crystals you have in the lower part here you have what we call a bayonet connection and here in the upper part we have cooling and so on so when we want to replace when this cartridge is full of impurities you have the possibility to here to install a casque ok and extract the cartridge replay and extract the extract the lower part store it before treatment for example cleaning in a cleaning pit install a new one empty of impurities so it means that when you have to replace you don't replace the whole system you have just to replace this cartridge we did that several times in Super Phoenix and it was very efficient even during the large pollution so it was a very good success but this system you can imagine that is not so easy to design because the diameter is 1.2 meters about and you are inside sodium at 550 and in the center you want to have 110 so it means that you need to have a very good insulation ok but it was ok expensive yeah the component is expensive but the cartridge after that is not expensive because you have no constraints even assuming that you have a leak you are inside the primary circuit so even if you have a small leak of sodium due to let's say a crack in a weld it's not really a big issue but yes it's expensive but for several people it was a big surprise to have no problems with this component ok here you have so you can see the shape here this is the upper part ok you can see the hole in which we insert the cartridge and in the lower part we have here you can see an electromagnetic pump an electromagnetic pump which circulates the sodium inside the length is 12 meters ok because we are 3 meters through the slab a particularity of this reactor just to recall something which is obvious for us but not for everybody in sodium fast reactors you can work ok you can you can see here some guy and personally I was several times because I was in charge of this component operation of this component and you can notice that you can work even in the reactor 100% of operation in this area so this is a peculiarity of this reactors liquid metal reactors ok but sometimes we have external loop here you have the secondary loop of superphénix in superphénix we have four loops four loops and one of them four loops you have here two intermediate exchanger we circulate the sodium and the sodium is sent to the steam generator here for each loop we have four loops each steam generator unit is 750 thermal megawatt ok so it's a huge monolithic we call that monolithic steam generator because there is another option to have modular steam generators it means several modules which are maybe more comfortable in case of an issue you can replace you can shut down one module replace it but without stopping the reactor fortunately in superphénix we didn't face problem like that this is the reason why the same option was also used for astrid project and here you can see the sodium at the outlet of the steam generator two pipes coming back to the pump the pump is installed on the cold leg of the system and sorry one pipe and then after the pump you go to two pipes going to the two intermediate exchanger for each loop we have two intermediate exchanger and one steam generator and here you can see here the cold trap so the cold trap is not immersed you have a circulation of sodium send to a cold trap a cold trap is here is what I have shown to you previously here you can see in a 3D view where you can see yellow the heater red the intermediate internal exchanger here the cooling area in green the coil with oil and in the jacket you have a sodium potassium and so on just different view one issue with this cold trap is that you can notice that we have deposits here but the problem with this concept is that which was designed in the mid of 70s is that you have deposits here so you have a variation of pressure drop inside and so very quickly you have a not very quickly but after some loading you have a circulation of sodium in the upper part where there is the lowest pressure drop ok and so you can have a circulation and we have observed in Phoenix after 4-5 years in fact the sodium go through here but is cooled by because the sodium is conducting thermal conducting media and here you have at the bottom when you enter in the inner pipe conditions to have a plugging ok so it means that at the end of the life of the cold trap we observe some plugging and so the feeling rate was not so good and second point is that the efficiency of the cold trap it's impossible in such a system to reach an efficiency of one ok you remember the description it's not possible to reach a minimum value because a part of the sodium go through the mesh where it is not super saturated ok so we have a profile you have a evolution of the profile of temperature evolution of the profile of circulation so you can understand that this the modeling of this cold trap is not so easy because an heat exchanger ok there are no changes in the flow ok and many for a pump the same but for this component because it's very specific you have not only heat transfer but we have also mass transfer ok ok yes it's just a demonstration of the efficiency a part of the flow rate is going in the upper part and so the overall efficiency here cannot be one by as I told you but now remember the kinetics ok what we have seen but for oxide hydride temperature is important because of the nucleation the first point and for the oxide the nucleation is as a lower kinetic because of the limiting step of diffusion so it means that it's generally good to have here you can reach in this trap you reach the cold point temperature here but we have an additional area at the cold point temperature where we have good conditions to give more time for the oxide to nucleate so it means that in this concept we have two zones one for the hydride ok without without mesh except on the wall we have some systems in order to the cake of sodium hydride in order to avoid suddenly to drop ok the cake in the lower part we have some elements to fix to fix the ok so we call this concept Psycos PE4 trap in fact in French separating impurities by crystallization of hydride and oxide of sodium ok it was the title of a movie horror movie during this period designed for super phoenix main advantage efficiency of one for no sodium hydride and for sodium oxide flow rate minimized to limit the heat loss in the circuit because when you have traps you cool down you lose some heat ok it's not large nevertheless it can be a parameter large capacity history was patented and then optimized sized by industry which was in fact sized for super phoenix and test in experimental loop and installed on the on the super phoenix during the two last years of the reactor and size also more recently for a large facility we anticipate to support Keops project ok I don't go in detail it's an explanation for EFR EFR project European fast reactor super phoenix was Italian, French and Deutsch, Belgium, Netherlands association après that our colleagues from UK came in the EFR project ok so they were I would say mostly four partners including UK and we have developed a new concept in order to improve again the loading capacity and to keep efficiency of one for sodium hydride and sodium oxide here you can see this is not real but it is design of the mock-up you can see there are some plates with some kind of I would say wires metallic wires to support the nucleation to promote at the beginning it's the circulation you go through the plates ok and when you have deposits slowly you have circulation around the plates ok like that mechanical engineering systems like for example distillation called pyramid here you have the evolution of the secondary culture in France so phoenix here super phoenix in phoenix we have not this exchanger in the upper part but it was not a good concept and the the cooling is inside the coil it's not good because if you have an ingress of oil you produce hydrogen and for the operator it can be a sodium water reaction so it's not really so for super phoenix they decided to have an external jacket with coil super phoenix second version and for EFR for EFR gas gas cooled pyramid the psychos and pyramid have the same height and same diameter in fact in the idea of EDF it was intended that after this first version we will use EFR for the secondary loops air cooled in order to avoid circulation of oil everywhere and so on to simplify the operation unfortunately as you know due to political decision in 1997 to shut down super phoenix ok so it's my big regret but story ok other cold traps just I show you in Russia in India sometimes we are not the same basis experimental basis to design the cold traps so here you sorry you can see that we have different geometries here you can see bf5 bn350 in Kazakhstan bn600 and so on so I don't want to go in detail but it was forcing last year to have a benchmark on the design of a cold trap so it was it is a task we hope to perform it but it means that we have different geometries ok and if you look in literature you will find a lot of designs ok but I think that it's quite interesting the design of this apparatus here you have a cold trap in India in fact if you look it's a mix of pyramid ok in the upper part ok and in the lower part of psychos and effectively if you look the paper if you look the paper produced by Ijeka colleagues in fact they have decided to have a mix of the two concepts with reference to pyramid and psychos ok qualification strategy this is an important topic qualification what we did in the past I think it's very important for you is we have we have a definition of the main requirements design of the one scale cold trap for the reactor we have identification of similitude rules it can be the heat transfer the residence time in each zone mesh density and then you have a downscaling of the mock up manufacturing of the mock up loading of the ancillary system you need to have an ancillary cold trap for example to distribute a sodium with a given pollution at the entrance of the mock up and then you operate in various conditions and you check the performance of your cold trap you can imagine that it takes a lot of time the qualification of the cold trap so today we have decided to have another tool we have a large progress in CFD we have also a lot of feedback we have the kinetics also and we have introduced that in a new code called Anais and Anais means advanced model for sodium sodium integrated purification systems it's also the name of my daughter it's not and and so we have developed this model and it's not so easy to model that why because in such a system you have flows you have a thermal exchange you have to model also mass transfer and you have also to take into account the deposits in terms of let's say geometry variation in fact the mesh of the calculations is not the same at the beginning of the life of the cold trap at the end so there is a permanent adaptation of the mesh for the calculation of the flow so it's quite complex but we did we take into account two types of crystallization in the mesh ok, this time and also on the cold walls I don't go in details you will see but after that we have made simulation of experimental data we got in the past ok we have all the results for the qualification of this component and we also model also in the upper part which was not the easiest part you have deposits for the hydride on the cold wall here you can see a view on the deposits you can see hydride deposits on the wall what is important to notice is that is a porous media discrystal so you don't have a significant variation of the performance of the heat exchanger because the sodium is you have a very high thermal conduction and so you can have deposits up to 8-10 cm without a significant variation of the cooling of the performance and so we have a very detailed model of this area for example here we have deposits we have what we call diffuse interface when you have an ion of hydrogen it diffuses through here and of course you promote the nucleation the growth first and also the nucleation on the external surface on crystals on crystals the basis for nucleation ok we have a propagation of the front of deposits so it's a ok it works and we describe the evolution of the deposits with time ok and we are able also to have both simulation of the hydrogen and oxygen in the lower part so now the purification here you have the purification system design methodology what is important identification of the impurities to be removed assessment of the source as the production rate continuous or discontinuous assessment of the potential things it means that in a reactor you have a cold trap but you have also corrosion ok so oxygen can be trapped also by corrosion so you need to take into account this phenomena induction specification on the removal rate evaluation of the amount of impurities to be removed from the circuit the amount you have to decide how long you want to operate the cold trap before to replace all the cold trap or only the cartridge in case of integrated purification system simulation with anise code in order to estimate if they are not so placed deposits and plugging too early plugging in some parts because you have a bad design somewhere ok same criteria how to control the feeling rate mass balance monitoring of the pressure drop of the pump if you increase the plugging you need to increase the intensity of the electromagnetic pump in order to compensate the increase of the pressure drop visual control by endoscopy sometimes we did that in order just for experimental and research purpose to see directly on the secondary trap what was really the deposits we did that and the last one is neutron transmission measurement we have developed a very specific notronic technique to control the deposits inside in fact here just I show you here you can see the cold trap in the center there is a symbol in which we have a possibility to send a fast neutron source we use generally californium 252 and when you have here the it's transported by pneumatic means and in front of we have a counter idiom 3 counter you count the neutrons of course at the beginning when the trap is new the neutrons are not thermalized by hydrogen but when you increase the loading with hydrogen we have an impact the neutron distribution the neutron the flux of neutron ok yes just you will see the synthesis and the storage casque and the neutron counter here so it means that you can control by moving the source and the counter to measure a profile of impurities so we did that on phoenix and super phoenix also here you can see the profile of impurities so initial is of course you have this is the reference you measure and you have a profile and so on so you can here on this concept which was not a reference concept it was a prototype with I would say I don't explain why but it was a bad design here we are notice that at the bottom here the cooler was inside and so it was a good place to have deposits of hydride but when they when we use this trap we have noticed that we are pretty sure that we will have some problems in the lower part with the deposition of hydride because as I told you hydride is able to deposit not in a mesh but in a cold wall and effectively you can see that at this level we can notice a large increase of the deposits and it was the reason why very quickly we decided to remove this trap and to set up new traps ok hot trapping hot trapping is when you have a very small volume of sodium to purify we developed that for an internal loop we never installed in Phoenix it was intended to have a specific sodium loop inside the core in order to test some materials yeah and so for that cold trap with yes sorry this experiment important control of oxygen because if you have samples and suddenly you have an increase due to the loss of cooling it's not a good situation with regards to the specimens material so we decided to have a hot trap because you don't need to cool the sodium and we have we have selected this composition we made a screening of different composition titanium alloy to trap the oxygen here we have obtained the kinetics we have a publication on that and we have the capacity so it's for small small amount of sodium in very specific cases for trapping activated corrosion products there was some developments in the past initially in Germany and EBR2 and with nickel foils you can trap these activated corrosion products idea of EBR in the US of EBR2 is to equip some fuel assemblies with these nickel foils in order to trap to trap these impurities activated corrosion products as you know we removed the fuel assemblies periodically so in fact it's not really an issue to extract this pollution to trap the cesium cesium you can trap with carbon traps it was used again in EBR2 but also in Rapsody in Rapsody to purify the sodium in some other facilities also maybe it was well known process and it's a solid foam solid foam of carbon you trap by absorption absorption sorry the cesium and it's very efficient also to decontaminate we are going to do this operation for Phoenix we are going to start decontamination of sodium prior to treatment of the sodium with NOAA process instrumentation I spoke about the pilgrimage oxygen meter oxygen meter there was a lot of developments in the 70s 80s on oxygen meters here you can see one which was really well design and very efficient it was Arwell oxygen meter what is an electrochemical oxygen meter you measure a difference of potential between a reference between a reference in this case it can be inside we have indium indium oxide in the lower part you have an electrical wire made for example of molybdenum when the sodium enters there is a difference of potential because the concentration in the sodium is there is some variations so you have a variation of the signal here you have so it's fixed in the pipe in the upper part you have some copper exchanger in order to freeze the sodium so you have to ensure the tightness of the system you freeze part of sodium in order to avoid extraction so it works very well so you have here electrolyte ok here you have a calibration curve it means that it's very accurate and it's possible to you need to calibrate these apparatus before to install on the facility but you have a very good stability in fact it was very efficient hydrogen meter the principle is you have a membrane nickel membrane on the other side you have vacuum you pump and hydrogen permeates from the sodium to the vacuum system and on the other side you have a mass spectrometer to measure the hydrogen this is one system there is another system electrochemical hydrogen meter there was some developments in the 80s and it was developed also by India and I think also by Russia it's maybe simpler in order to compare with our system we install it in Phoenix reactor 12 years ago and compare the signals of conventional hydrogen meter by permission electrochemical and you can notice here this is on the lower part the electrochemical hydrogen meter you can see that it is you detect variation of hydrogen same time no delay very good comparison on the two thank you for your attention thank you very much Christian so we have one minute for the questions I will ask one about SMR because as you develop the SMR sodium cooled SMR this is a device like big 8 meters tall 2 meters diameter for the culture and then if you want to do it for this purification for the sodium cooled SMRs what is the solution do you also need this you mean for example internal integrated system no just what ideas how can you do it called traps for the small for small modern reactors but I think that I think it will be an optimization but I think that internal integrated option is a good option I think we are able to one point is the diameter of the systems generally so if we integrate the cold trap we need to reduce the diameter so we think it is possible because I did that for a project called SPX2 a long time ago so I think it is possible and effectively the idea is to avoid one of the big advantage I didn't mention that but if you have integrated the cold trap you have no circulation of active sodium out of the main vessel so it's a big advantage in terms of safety assessment and so I think that it seems to me that integrated system would be more convenient for these small modern reactors because if you start to have ancillary systems we see ok thank you any other questions thank you for presentation Dr. Leig my question about the maintenance time of the cold trap if I understand correctly you say that because of the high pollution the reactor shutdown and the maintenance what's the maintenance time for for a cartridge changing ok I understand so for a cartridge for example in Super Phoenix it was maximum 1 week to replace the two cartridges it can be done during the handling fuel assemblies fuel assemblies handling in parallel but you need to know what is the ratio of the feeling of your cold trap so you need to follow the evolution of the cold trap in order to anticipate you can say for example next stop shutdown normal shutdown of the reactor for the fuel assemblies the management you can foresee also in parallel way 1 week dedicated to the removal of the cartridge if it is another system if it is an external loop you can have of course extract cut the cold traps and you have a spare cold trap and what to do on the main cold trap we have 2 options or you are very rich ok you can wait for the decommissioning ok clearly and our second option is also to what we call regeneration there is some process to regenerate we have developed one for EFR we have a regeneration process where we dissolve again the impurities and we separate oxygen and hydrogen triciated hydrogen because we have tricium and we manage the tricium hydrogen triciated for example by oxidation and after that we reinstall the external cold trap generally we don't have seals and joints generally we prefer to have wedding in order to avoid any risk of leak I know if I know it correctly in a regular refueling time this system also maintenance in a regular refueling times but in this time is a particular time that it is occur for the cartridge it was every 2 years every 2 years and for external cold traps for the first generation which was changed it was every 4 5 years and the new generation of cold traps it was anticipated particularly the last one every 12 or 15 years thank you and more question the primary and secondary sodium purification has a separate cold trap or one cold trap for 2 loops for primary and secondary it's different different separate because for one reason the primary you can have activated corrosion products on the secondary the only contaminant is tritium tritium of course we have to deal with but it's a very small amount and it's not the primary is loaded mainly with oxide and the secondary with hydride in a steady state operation ok thank you thank you yeah please thank you Dr Lachey on the basic side why is tritium produced on the secondary because of absorption of notion no it's not in the secondary in the primary but during my next presentation I will explain this point ok again thank you very much so we are nearly in time only 5 minutes late and let's have a coffee break and start at 10 10.50 exactly and I will put this