Home Energy Physics Nuclear Power Electricity Climate Change Lighting Control Contacts Links



By Charles Rhodes, P.Eng., Ph.D.

This web page outlines the logic behind sustainable advanced reactor technologies. It is shown that for dry land based electricity generation liquid sodium cooled Fast Neutron Reactors (FNRs) operated at 500 degrees C with Fe-Cr-Mo fuel tubes offer the advantages of both low material cost and low fuel reprocessing cost. Marine reactor applications for which liquid sodium is an unacceptable reactor coolant can in principle be met using thermal neutron reactors operated at 650 degrees C with an isotopically pure molten salt coolant. However, there remain unresolved material issues for reactors with molten salt coolants.

Modern advanced nuclear reactors used for public electricity generation must simultaneously meet multiple objectives including:
1) Fuel sustainability;
2) Nuclear waste disposal;
3) Certain avoidance of prompt neutron criticality;
4) Capable of sustained reactor fleet growth;
5) Minimal maintenance cost;
6) Long equipment operating life;
7) Efficient electricity generation;
8) Safe in urban siting;
9) Modest cost;
10) Power control.

There are many published paper reactor concepts that for various reasons do not meet all of the aforementioned objectives.

A major issue in the choice of an advanced nuclear reactor technology is long term fuel sustainability. The only fuel sustainable fission reactors are breeder reactors. There are two fuel breeding cycles. One fuel cycle uses fast neutrons to convert U-238 into Pu-239 and then fissions the Pu-239. The other fuel cycle uses thermal or fast neutrons to convert Th-232 into U-233 and then fissions the U-233. However, the Th-232 based reactors rely on fissile start fuel supplied by the U-238 based fuel cycle for reactor fleet growth and rely on surplus neutrons from the U-238 based fuel cycle for nuclear waste disposal. The Th-232 fuel cycle also has complications related to the high thermal neutron absorption cross section of Pa-233 and related to formation of U-232 which has a decay sequence with a hard gamma emission. Since at this time we urgently need to both grow the nuclear reactor fleet and dispose of nuclear waste this web page focuses on U-238 fuelled fast neutron breeder reactors.

Breeder reactors must be large enough to minimize neutron leakage from their fuel surrounding blankets. When there is a sufficiently large fleet of U-238 based breeder reactors to provide the fissile start fuel required by Th-232 fuel cycle reactors humans can pursue breeding of Th-232 into U-233.

As set out above, in order to achieve both fuel sustainability and nuclear waste disposal it is necessary to use fast neutron breeding of U-238 into Pu-239 to build a reactor fleet. However, fast neutrons are very destructive. Over time a fast neutron flux causes lattice dislocations, material swelling, embrittlement and chemical changes. The solution to this problem is to surround the reactor fuel assembly with a sufficient thickness of a stable liquid coolant to fully absorb leakage neutrons before the neutrons can reach reactor structural materials or the intermediate heat exchanger. Hence the fuel assembly should be contained in the center of a large liquid coolant pool and the heat exchangers should be located at the edge of the liquid coolant pool. Periodically the entire fuel assembly should be replaced and its materials recycled.

A pool type liquid metal cooled or molten salt cooled reactor inherently operates at a low pressure and provides a substantial thermal mass which eliminates the pressure vessel risks inherent in reactors which contain high pressure super heated water. In a pool type reactor the coolant boiling point is far above the coolant's normal working temperature. The pool surface temperature is typically in the range 475 C to 650 C which is sufficient for efficient electricity generation.

The concept of dissolving nuclear fuel in primary coolant liquid salt generally proves unworkable in a pool type power FNR due to the coolant containment walls and heat exchanger being exposed to neutron damage, due to potentially dangerous prompt neutron criticality in an earthquake, due to extreme tangential stresses on the containment pool walls during an earthquake and due to fission products corroding and/or depositing on heat exchange surfaces. Furthermore, the radio active fission product Cs-137 will evaporate from the coolant pool and condense all over everything in the reactor space, making future heat exchanger replacement and other maintenance work somewhere between extremely difficult and impossible.

In order to achieve a long reactor working life and maintain a high reactor thermal output it is essential to prevent fission products migrating from the fuel to the primary coolant enclosure walls and depositing on the heat exchange surfaces. Meeting this requirement requires that the fuel be contained in sealed metal fuel tubes. The fuel tube material must have a high melting point and should have a small neutron absorption cross section. Over time the fuel tube material will be damaged by fast neutrons, so the fuel tubes must be replaced when the fuel is reprocessed. To be economical the fuel tube material should be primarily iron based. Ideally the fuel tube material should be recycled.

A key issue with sealed metallic fuel tubes is that they contain some metallic sodium. This metallic sodium serves three important functions:
a) Chemically combining with fission product gases to reduce the gas pressure inside the fuel tube.
b)Chemically combining with corrosive fission products such as F, Cl, I, Br to prevent internal corrosion of the fuel tube.
c) Providing good thermal contact between the solid U-Pu-Zr fuel alloy and the fuel tube.

Note that it is impossible to maintain metallic Na inside a fuel tube containing oxide fuel due to ongoing oxide reduction. This use of Na for fission product gas absorption only works for metallic fuel.

Moltex contemplates use of vented fuel tubes and a molten salt fuel. This scheme has several potential problems:
a) Requires highly enriched fuel to maintain reactor core zone criticality with a fast neutron spectrum;
b) Internal corrosion of the fuel tubes which reduces fuel tube life;
c) Heat flux degradation due to fission product corrosion and/or deposition on both fuel tubes and heat exchanger tubes;
d) Long term corrosion of the primary coolant containment walls;
e) Primary coolant becomes highly radioactive;
f) The radio active fission product Cs-137 evaporates from the coolant pool and condenses on every cooler surface in the coolant pool enclosure, making future heat exchanger replacement difficult to impossible.

Fast neutrons have typical energies of 2,000,000 eV. Crystal lattice binding eneries are of the order of 1 eV. The mass of a metal lattice atom is typically 50X the mass of a neutron. Hence, in an elastic collision a fast neutron can easily transfer 2,000,000 eV / 50 = 40,000 eV to a metal lattice atom. This transferred energy is sufficient to severely disrupt the metal crystal lattice. When a material is exposed to a fast neutron flux the resulting lattice dislocations tend to make the material less dense. Typically a fast neutron flux will gradually convert a metal FCC (face centered cubic) crystal lattice into a less dense metal BCC (body centered cubic) crystal lattice, which phase change causes material swelling.

In order to minimize fast neutron induced fuel tube material swelling either the fuel tube's fast neutron cross section should be minimized so that the material swelling is tolerable or the fuel tube material should initially already have a BCC crystal structure. If the fuel tube material initially has a dense hexagonal close packed (HCP) metal crystal structure or a FCC crystal structure it will start swelling immediately it is exposed to a fast neutron flux.

A steel alloy which is normally BCC at temperatures below 475 degrees C is Fe + 12% Cr + 1% Mo. This alloy is known HT-9. Absent a fast neutron flux above 475 degrees C the crystal structure of HT-9 spontaneously changes from BCC to FCC. As the material is exposed to a fast neutron flux it experiences lattice dislocations. Below 650 degrees C the material does not significantly swell but in a fast neutron flux the material becomes very brittle. At 650 degrees C the material anneals releaving the brittleness, which seems to indicate that the lattice dislocations disappear.

If it is desired to operate FNRs to 15% fuel burnup before fuel reprocessing it is essential to prevent fuel tube working life limiting fuel tube swelling. Hence it is necessary to live with some fuel tube enbrittlement.

If the reactor is operated at 650 degrees C either liquid sodium or molten salt coolant can be used and the fuel tube enbrittlement is relived. However, at that temperature there are potential problems with Pu melting.

A possible alternative fuel tube material is zirconium. It has a HCP lattice and will start swelling immediately it is placed in a fast neutron flux but the swelling rate will be relatively low due to its relatively small neutron cross section. One advantage of Zr as a fuel tube material is that its higher permitted operating temperature potentially allows use of a molten salt as a primary coolant in place of liquid sodium.

Zirconium has a much smaller fast neutron absorption cross section than iron or chromium. However, at temperatures less tha 863 degrees C zirconium has a hexagonal close packed crystal structure. Hence, although it has about a 14X smaller neutron absorption cross section than Fe and Cr, Zr will start to slowly swell as soon as it is exposed to fast neutrons. Thus, unless a very high operating temperature is essential, Zr offers few tangible benefits over Fe-Cr-Mo and Zr is much more expensive.

Sigma Zr = 0.184 b
Sigma Fe = 2.56 b
Sigma Cr = 3.1 b

In order to transport heat by natural convection from the fuel tubes to the heat exchanger the primary coolant melting point should be at least 150 degrees C below the maximum coolant operating temperature. If the maximum coolant operating temperature is 500 degrees C this constraint eliminates molten salt coolants but is compatible with liquid sodium. Thus for compatibility with Fe-Cr fuel tubes operated at up to 500 degrees C the primary coolant should have a melting point less than 350 degrees C. The primary coolant boiling point must be far above 500 degrees C. Simply on the basis of melting and boiling points the possible elemental liquid primary coolant choices are bismuth (271.4 C), tin, lead, mercury, sulfur and sodium. Bismuth and tin when neutron activated form undesirable toxic long lived radio isotopes. Sulfur is too chemically corrosive. Mercury is dangerously toxic and has a relatively high vapor pressure in the contemplated operating temperature range. Sodium is chemically compatible with steel and has excellent nuclear properties. Lead is very dense, is less chemically compatible with steel and has inferior nuclear properties.

If the contemplated operating temperature is 650 degrees C in principle various salt mixtures might be used for the primary coolant. However, to minimize corrosion and nuclear waste accumulation the salts with low melting points rely on isotopically pure Li and Cl components, both of which are expensive. A further problem is that when Li-7 absorbs a neutron it decays to form He-4. In so doing it releases an agressive Cl or F atom from the salt that corrodes the containment wall. Absent the Li the salt melting point rises which triggers other forms of corrosion.

A conservative approach to this whole issue is to build a liquid sodium cooled reactor intended for up to 525 degree C operation. While this reactor is being designed and built the issue of the optimum fuel tube allloy and its preferred operating temperature will likely be better resolved. The danger of committing to a molten salt cooled reactor at this time is lack of certainty about the existence of sufficiently corrosion resistant materials for use in molten salt cooled reactors. Since development time is of the essence the more conservative path to pursue is liquid sodium cooling because we know it will work, whereas molten salt cooling has multiple technical uncertainties.

For power reactors, from a material cost perspective, sodium as the primary coolant has the advantage that it is inherently inexpensive. The disadvantages of sodium are almost entirely related to its potential chemical interaction with air and water and its low specific heat as compared to water.

When a very fast neutron hits a sodium atom nucleus a stable fluorine atom and a He-4 atom are produced. The fluorine atom immediately chemically combines with another sodium atom to produce a NaF sludge which sinks to the bottom of the primary sodium pool. The He-4 is a harmless off gas. When a thermal neutron is absorbed by a sodium atom after about 15 hours a stable magnesium atom is produced. The magnesium is more dense than liquid sodium and again forms a pool bottom sludge.

In the event of a fuel tube wall failure the use of sodium both inside and outside the fuel tube minimizes the flow of fission products from inside the fuel tube to outside the fuel tube.

These simple properties indicate that the primary coolant should be metallic sodium.

The choice of sodium as the primary coolant forces the thickness of the protective layer of sodium between the fuel assembly and other objects to be about 3 m. Then the heat exchangers and pool walls so protected have very long operating lives. Sodium has a relatively large thermal coefficient of expansion which contributes to stable reactor primary coolant temperature control.

In order to achieve breeding the fuel assembly consists of a core zone surrounded by a blanket zone. To enable reactor power control by coolant thermal expansion the core zone is pancake shaped, being much wider and longer than it is thick. To compensate for the effect of fuel aging on the reactor temperature setpoint the fuel geometry is slowly changed by changing the vertical position of some of fuel bundles with respect to their neighbouring fuel bundles. This temperature setpoint modification method changes the average fissile fuel density in the reactor core zone. This arrangement is made safe against gravity by inserting mobile fuel bundles from the bottom rather than from the top.

To achieve high breeding efficiency the reactor core is surrounded by a thick U-238 blanket. Through the use of stacked U-238 fuel rods the blanket is divided into three zones, the inner blanket, the middle blanket and the outer blanket. After each fuel cycle the inner blanket fuel rods are reprocessed, the middle blanket fuel rods are moved to become the inner blanket rods, the outer blanket fuel rods are moved to become middle blanket fuel rods and the outer blanket is formed using new U-238 fuel rods.

A practical power FNR contains as many as 1369 fuel bundles, each of which contains about 3000 fuel rods. The fuel bundles are periodically replaced. To make FNRs economic the entire fuel reprocessing and fuel bundle fabrication procedure must be automated.

Thus the basic configuration of an economic FNR is a liquid sodium pool containing an assembly of vertical fuel tubes in the center of the pool and heat exchangers around the perimeter of the pool. The fuel tubes are in bundles. Metallic fuel rods are stacked in the fuel tubes so as to realize a pancake shaped reactor core zone and three surrounding blanket zones. The reactor temperature setpoint is adjusted by moving mobile fuel bundles up or down with respect to their adjacent fixed fuel bundles to change the reactor core zone geometry.

At this time the most satisfactory FNR fuel tube material is Ht-9 (Fe-Cr-Mo). If liquid sodium coolant is used the reactor should be operated at about 500 degrees C. If the chemical properties of sodium are unacceptable a molten salt coolant can be used for 650 degree C operation, but there will be corrosion problems. Suitable molten salt coolants are generally expensive due to required isotopic separations. A typical molten salt is 42% Li-7 Cl-37 - 58% K Cl-7 which has a melting point of 352 deg C. Note that expensive isotopic separations are required to obtain the Li-7 and Cl-37. Note that molten salts have high viscosities at temperatures within 100 degrees C of their melting point, so the practical operating temperature range of a molten salt cooled reactor is about 500 degree C to 650 degree C.

For dry land based bulk electricity generation due to material cost economy a liquid sodium cooled power reactor with Fe-Cr-Mo fuel tubes is the preferred technology. A liquid sodium cooled reactor must be sited at a sufficient elevation above its immediate surroundings that it will never be exposed to flood water. Marine safety requirements will likely justify the additional costs of a molten salt cooled FNR.

A liquid sodium cooled FNR operated at 500 degrees C can be operated to 15% fuel burnup.

The most cost effective way to use a nuclear reactor is for base load electricity generation. However, as fossil fuel generation is eliminated from electricity systems it is increasingly necessary for nuclear reactors to provide real time power control. The relevant issues are discussed in the document: Non-baseload Operation in Nuclear Power Plants: Load Following and Frequency Control Modes of Flexible Operation

This web page last updated April 1, 2020

Home Energy Physics Nuclear Power Electricity Climate Change Lighting Control Contacts Links