Patent Application
20230197299
The present invention relates to fission reactors. In particular the present invention relates to reactor designs and methods of operation in order to control the formation of bubbles within the fuel salt of a molten salt reactor.
Nuclear reactors using molten salts as fuel (MSR's) have been known since the 1960's when the US Molten Salt Reactor experiment was successfully run. Many designs for molten salt reactors have been conceived since that time. Such reactors fall into two classes.
In the first class, pumped MSR's, the molten salt fuel is actively pumped between a reaction chamber, where the fuel enters a critical state and generates fission heat, and a heat exchanger where the heat is transferred to another fluid, often another molten salt without fissile elements (a “coolant salt”), and used to generate power.
In the second class, static molten salt reactors, the molten salt fuel sits within fuel tubes, often formed into fuel tube assemblies, and moves within the tubes only through natural convection, with the fuel remaining within the reactor core and not being pumped outside that core. A second fluid (e.g. a coolant salt) flows past the fuel tubes and removes heat from the core. The fuel salt remains in the core throughout its operational life and is refreshed by removing a spent fuel assembly and replacing it with a fresh one—essentially as is the case with fuel assemblies containing solid fuel elements. This second class of MSR is described in GB2508537 and equivalents.
Molten salt reactors inevitably produce noble gasses as fission products and those gasses have low solubility in the molten salt fuel. It is essential for the safety of the reactors that the movement of the noble gasses out of the active reactor core is highly predictable since one of those gasses is Xenon 135 which is the strongest neutron absorber known and therefore has major influence of the reactivity of the core.
In the first class of pumped molten salt reactors those gasses can be removed in a variety of ways including sparging the fuel salt with Helium.
The challenge of managing noble gasses in the second class of molten salt reactors is greater. The gasses can diffuse out of the molten salt fuel into the gas space above the fuel where they will have only small reactivity impacts since they are substantially outside the reactor core. However, it is also possible that they will form bubbles on the inner surface of the fuel tubes. Such bubbles could be triggered to detach from the tube by a physical shock or other effects with a large number of bubbles simultaneously rising up the fuel salt into the gas space and causing a major increase in core reactivity.
This would be an unacceptable safety vulnerability and there is therefore a need to ensure that bubble formation in the fuel tubes cannot occur.
According to a first aspect of the invention, there is provided a molten salt fission reactor. The reactor comprises a reactor core, which comprises a plurality of fuel tubes. Each fuel tube contains a fuel salt and a gas interface. The fuel salt is a molten salt of one or more fissile isotopes. The gas interface is a surface of the fuel salt in contact with a gas space during operation of the reactor. The reactor also comprises a fuel salt cooling system, which is configured to cool the fuel salt. The cooling system comprises a heat exchanger and a coolant tank. The coolant tank contains a coolant liquid in which the fuel tubes are at least partially immersed. The heat exchanger is for extracting heat from the coolant liquid. The fuel salt cooling system is configured such that during operation of the reactor, for all points within the fuel salt within each fuel tube except at the respective gas interface:
where:
Accordingly, the solubility of Helium is lowest at the gas interface.
The fuel salt cooling system in configured such that during operation of the reactor the solubility of a noble gas (e.g. Helium as described above, or any other noble gas from Helium to Xenon) in the fuel salt close to the gas interface is lower than its solubility elsewhere in the fuel tube. This ensures that the noble gas does not reach a saturating concentration other than in this region close to the gas interface and bubbles of gas cannot therefore form.
The solubility of noble gas in the fuel salt is a function of:
The latter two factors are most amenable to control.
If the temperature of the fuel salt at the top of the tube can be maintained at a lower level than the temperature lower down the fuel column then gas bubbles cannot form and gas will exit the fuel salt by diffusion across the gas interface. That low temperature can be achieved by one of more of the following mechanisms:
However, the pressure effect is also highly relevant in such systems. Fuel salt is a dense liquid and the hydrostatic pressure of the liquid column means that, at constant temperature, the solubility of noble gasses increases at lower levels within the fuel salt column.
It is therefore possible to achieve the objective of having minimum gas solubility at the upper salt/gas interface even though fuel salt temperature is somewhat lower further down the salt column.
The balance of the temperature and pressure effects dictates the safe operational envelope for the nuclear reactor.
According to a second aspect of the invention, there is provided a method of operating a molten salt fission reactor according to the first aspect. The temperature of the fuel salt is maintained such that that during operation of the reactor, for all points within the fuel salt within each fuel tube except at the respective gas interface:
where:
Accordingly, the solubility of Helium within the fuel salt is lowest at the gas interface.
This may be done equivalently for noble gases other than Helium.
In developing the safety case for the Static Molten Salt Reactor described in GB2508537 we have discovered that there are serious potential hazards associated with noble gas behaviour in molten salt fuelled reactors that have not been adequately addressed in current designs.
Potential mechanisms through which significant transients in reactor power could occur centre around the presence of gas bubbles within the reactor core, particularly within the critical region (i.e. the region in which the density of fissile isotopes is sufficient for a self-sustaining nuclear reaction to occur, during operation). Provided the concentration of gas dissolved in the fuel remains below its saturating concentration, bubbles cannot form and it is straightforward to calculate the impact on the reactor physics of the dissolved gasses. If bubbles form however, far more complex phenomena can ensue.
Xenon gas dispersed in solution in the fuel salt will have greater neutron absorbing effect than the same amount of gas in a bubble. This is because Xenon 135 is such a strong absorber that Xenon in the centre of a bubble containing Xe135 will be substantially shielded from neutrons by the bubble itself. Formation of a bubble from a supersaturated solution of gas in the fuel salt could therefore substantially reduce neutron absorption and hence result in an unwanted and uncontrolled increase in the reactivity of the core which will lead to a rapid and potentially unmanaged increase in reactor power.
Similar concerns arise with the other noble gases, as any gas bubbles will displace fuel. This can potentially lead to a large increase in reactivity if a significant volume of bubbles within the reactor are displaced at once, and rise to the surface from within the critical region of the fuel tube.
When previous data is looked at in light of these findings, a vital clue that problems may exist arose from the operation of the Molten Salt Reactor Experiment. In the report ORNL 4396 (p17) it was reported that apparently random blips in reactor power of around 10% occurred with significant frequency. Despite much investigation, a clear explanation of these power fluctuations was never reached. It was believed however that behaviour of gasses in the fuel salt circuit was responsible and an empirical solution that eliminated the blips by adjusting the pump pressure and gas sparging was achieved. The Molten Salt Reactor experiment operated at very low power of <10 MW. Commercial reactors will operate at radically higher power. Since the magnitude of the Xenon effect is proportional to the power this potentially means that a fluctuation that may only be an irritation at very lower power may become a serious hazard at higher powers. In the Molten Salt Reactor Experiment there was the capability of changing the Helium sparging rate to control the Xenon bubble issue. No such capability exists in molten salt reactors such as those described in GB2508537 where gas loss from the fuel salt can only occur by passive diffusion into the gas space above the fuel salt
The problem of supersaturation may be greatly exacerbated by the actual nuclear fission process in the fuel salt.
At a power of 100 kW/litre of fuel the number of fissions per second can be calculated thus:
A 10 um diameter bubble would have an approximately 25% probability of being impacted by any fission fragment produced within a 20 um diameter spherical volume of 4/3*3.14*1e−12=4e−12 litres. The bubble would therefore be impacted on the order of a thousand times per second.
The energy in such a fission fragment is orders of magnitude higher than the surface energy of the bubble and results in temperatures in the medium through which the particle passes in the 10's of thousands of degrees. It is likely therefore that it would result in the destruction of the bubble, with the gas content redissolving in the salt. A very high level of supersaturation of gas in the molten salt may therefore be maintained so long as fission continues. A feedback loop between the rate of fission and the reactivity of the reactor core may therefore establish and could readily lead to instability.
Other potential phenomena include deposition of bubbles of gas on solid surfaces within the reactor core. Such bubbles could accumulate for some time and be displaced from the solid surfaces by a shock wave, vibration or flow disturbance of the fuel salt. In that event a large amount of neutron absorbing material may exit the core in a short time period resulting in a sudden increase in core reactivity and potentially a damaging power surge. Furthermore the removal of void volume from the core, replaced by fissile fuel salt (i.e. increasing the average fuel salt density within the critical region), could lead to another increase in reactivity.
While these phenomena are of concern in all molten salt reactors, they are particularly concerning in the static molten salt reactor class because the fuel salt in such reactors flows at a relatively low speed under natural convection and bubbles are therefore more likely to form and accumulate on the inner surface of the fuel tubes than in reactors where the salt is pumped at relatively high velocity.
It is evident that it will be exceptionally difficult to ensure that no unexpected hazardous effects could occur in a molten salt reactor where bubbles are allowed to form within the critical region of the reactor core during operation. It is therefore highly desirable that the concentration of dissolved gasses in the molten salts in the core of such reactors be maintained at below their saturating concentrations, thereby making formation of bubbles within the core region impossible.
The solubility of noble gasses in molten salts increases with increasing temperature, the opposite of the behaviour of gasses in water (ORNL-2931 Reactor Chemistry Division annual progress report Jan. 31, 1960). The consequence of this is that where there is a gas/salt interface where the salt is at its highest temperature, the gas will dissolve in the salt. As the salt is cooled, the gas can therefore become supersaturated in the salt and be prone to forming bubbles.
Additionally, as fission gasses are continuously produced in the fuel salt, the concentration of these gasses in the fuel salt will rise until off-gassing occurs at a rate equal to production. In a well-mixed salt where the concentration is the same in all locations the off-gassing will occur where the solubility is lowest—if this is not the surface, then bubbles will come out of solution lower down in the fuel.
In the description below, we propose arranging the design of the molten salt reactor so that the salt in contact with the gas phase has a lower gas solubility than any other point in the fuel. Dissolved gas will therefore diffuse out the salt across the interface without formation of bubbles.
Most Molten Salt Reactor designs, including that described in GB2508537 take advantage of the increased buoyancy of molten salts as they heat to partially or completely drive the circulation of the molten salt. This necessarily requires the direction of salt flow to be upwards. Unfortunately however, this also means that the coolant salt temperature is highest at the top of the fuel tube which makes it difficult to avoid the fuel salt in that region being hotter and therefore having a higher solubility for noble gasses than further down the core, especially at the bottom of the tubes, where the coolant is at a lower temperature.
One way to lower the temperature of the fuel salt near the gas interface is therefore to reverse the direction of flow of the coolant salt to vertically downwards so the coolant salt temperature increases as it flows down through the core. This has the disadvantage of opposing the natural convection forces but ensures that the upper regions of the flow circuit are at the lower temperature.
Where reversal of coolant flow direction is not possible, for example in static molten salt reactors where coolant flow is by natural convection, it is still possible to ensure that the upper surface of the fuel salt is the point of lowest gas solubility in the fuel salt. This can be achieved by one or a combination of the following methods:
The relative temperatures of the fuel tube wall, the coolant salt at the bottom of the tube, and the coolant salt at the top of the tube will depend on the fuel tube dimensions, the flow rate of the coolant salt, the temperature of the coolant salt that enters the core, and the power density of the nuclear reaction. In particular, a model that has been found to be able to test the viability of solutions contains:
The model can be used in two ways—either assuming that the tubes generate heat uniformly along their length (up to the level of the top of the fuel salt) for a simpler model, or taking into account the vertical gradient of the energy production in the fuel tube for a more accurate model. For example, the simpler model may be used to identify target parameter values, which are then checked with the more accurate model or experimentation.
For a given power density, coolant salt inlet temperature, fuel tube length, fuel tube diameter, and annular thickness, the coolant outlet temperature and the temperature of the fuel tube at the top of the fuel salt can be calculated. The difference between the coolant outlet temperature and the temperature of the fuel tube cladding at the top of the fuel salt will be equal to the difference between the coolant inlet temperature and the temperature of the fuel tube cladding at the bottom. The coolant inlet temperature is a defined parameter, and can therefore be used to calculate the temperature of the cladding at the bottom of the fuel tube.
As a further example, by fixing all but two of the input variables in the model (e.g. fixing the power density, inlet temperature, and fuel tube length), graphs can be plotted of the coolant outlet temperature and cladding temperatures as dependent on the other two variables (in this case, the fuel pin diameter and annular thickness. An example of this is shown in
Other simulation, modelling or prototyping methods and techniques as known in the art may alternatively be used to determine the parameters required.
It will be appreciated that this analysis is only needed for simulation—for an actual reactor, the temperatures can simply be measured to determine whether they are in the correct relationship.
Optionally, neutron absorbing structures can be inserted inside or outside the fuel tube in any position from just below the surface of the fuel salt to part way up the gas space in the fuel tube so that fission is suppressed in the uppermost layers of the fuel salt.
The general equation for gas solubility in a fluid is given by Henry's law: ca=P*Hcp, where ca is the solubility in Mol/cc, P is the partial pressure of the gas at the surface in atmospheres (1 atm=101325 Pa), and Hcp is the Henry Solubility of the fluid. This can be made temperature dependent by substituting Hcp with
where H(T) is the updated henry's constant for temperature T, Ho is the henry's constant at reference temperature To, ΔH is the enthalpy of solution, and R is the gas constant for the gas involved. Note that the “H” in the variables ΔH and H are entirely different quantities, and should not be confused.
Any bubble surfaces that formed deep in the salt would be at the hydrostatic pressure of the salt at that depth. This means that the partial pressure of the gas in the bubble would be equal to the hydrostatic pressure (for a single gas). Thus, when calculating the saturation concentration in below-surface regions, the salt hydrostatic pressure at that point is used rather than the salt surface pressure.
As the fuel is 1800 mm deep and has an average density of 3181.7 kg/m3, the absolute pressure at the bottom of the pin is 1.561 Bar. The temperature of the salt at the bottom of the pin is 825.3° K, so the fuel there will saturate at the higher concentration of 2.0633*10−8 Mol/cc despite the lower temperature.
For the fuel at the bottom of the tube to saturate at the same gas concentration as the top, the temperature at the bottom would have to be equal to the temperature
where P1 and T1 are the pressure and temperature at the fuel surface, and P2 and T2 are the pressure and temperature at the bottom of the tube. This relationship holds true for any depth in the fuel. For the example shown in
Therefore, to ensure that the Xenon does not come out of solution except at the gas interface, the reactor is configured such that
It should be noted that the examples discussed earlier where T2>T1 will always satisfy this relationship, as the right hand side will always be less than T1.
A locus 702 of temperatures are shown in
As the solubility limits described by the Loci 702 and 703 are relative to the fuel surface, they are not affected by the constant H°.
In a well-mixed system where the fuel surface is saturated and cannot dissolve more gas, the loci show the temperatures the fuel would have to fall to in the rest of the tube to also become saturated.
While the above examples have focussed on Xenon as an exemplary gas of concern, it may be possible to design reactors which could tolerate limited formation of Xenon bubbles, but would not be able to tolerate the formation of bubbles of lower atomic number noble gases while still operating safely. For such reactors, the above analysis can be conducted, but with the gas constant R substituted for the gas constant of Helium, Neon, Argon, or Krypton, and the enthalpy of solution ΔH substituted for the enthalpy of solution of that gas with the fuel salt. The gas constant R (in J/kg ° K) is higher for lighter noble gasses, as is ΔH (in J/kg), though it rises at a slower pace. This means that any reactor which satisfies this criterion for a certain noble gas (e.g. Argon) also satisfies it for all noble gases with lower atomic number (e.g. Helium and Neon), as the value of R/ΔH will increase and thus the minimum temperature T2 will decrease.
For many combinations of noble gas and molten salt ΔH can be obtained from the literature. Where such a previous measurement is not available, ΔH can be experimentally determined by a process of:
Each of these steps will be considered in more detail below.
Solubility of the gas in the molten salt at a given temperature and pressure can be measured by any suitable means. One example is described in W. R. Grimes, N. V. Smith, and G. M. Watson, J. Phys. Chem. 62, 862 (1958), where the gas solubility in a salt can be measured by allowing a sample of salt to saturate with a pure stream of the test gas while maintaining conditions at the temperature and pressure of the desired measurement point. This salt sample can then be isolated and sparged with a different gas to remove the dissolved test gas. The level of test gas in the outlet stream can then be measured and the saturation concentration calculated. Note that Grimes et. al. uses the symbol K for the Henry's law constant Hcp.
The Henry's law constant Hcp is the gradient of a graph of solubility against pressure for a given temperature, or can be obtained from a single measurement by dividing the solubility measurement by the pressure (though as always, taking a gradient from multiple measurements will provide improved accuracy).
Once Hcp has been found for a set of temperatures, a plot of the temperature-dependent Henry's constant H will follow the Van't Hoff equation:
Where Ho is the value of H at a reference temperature To, and R is the gas constant of the gas. The reference temperature and corresponding reference value of Ho can be chosen as one of the measured points, and then the parameter ΔH can be varied to obtain a best fit (e.g. measured as a minima in the total squared error between the points and the curve). The value of ΔH which produces the best fit can then be used to determine the temperature relationship set out above for that particular gas/molten salt pair.
The above derivation is identical regardless of the units used (provided temperature is measured in a system such as Kelvin which treats absolute zero as 0). The solubility is typically expressed as moles per unit volume (often mol/cm3), Ho is typically expressed as moles per cubic centimetre per atmosphere (mol/cm3/atm) or equivalent units. ΔH is measured in Joules per kilogram (care should be taken with units of ΔH and R−ΔH is often given in cal/mol, so should be converted for use with SI values of R).
ΔH for a given gas/molten salt pair will be independent of temperature and pressure (provided the gas is a gas and the molten salt is a liquid). Therefore, the choice of temperature and pressure values used to determine ΔH should not affect the final result. However, suitable values for pressure would be, for example, the maximum and minimum operating pressure of the reactor, and their midpoint, or 0.5 atm, 1 atm, and 1.5 atm. Suitable values for temperature would be, for example, the maximum, and minimum operating temperature and their midpoint, or at 100, 200, and 300 degrees above the melting point of the molten salt.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2007517.2 | May 2020 | GB | national |
| 2010754.6 | Jul 2020 | GB | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2021/063373 | 5/19/2021 | WO |
