Home Energy Nuclear Electricity Climate Change Lighting Control Contacts Links



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

The purpose of this web page is to set out the various fuel processing and material recycling steps that are required to support a Fast Neutron power Reactor (FNR) operating under the Ottensmeyer Plan. Note that no material exits from the fuel treatment process until its radio toxicity is below the radio toxicity of natural uranium.

which is intended to be performed on a CANDU reactor site:

Consider 1000 gm of spent CANDU fuel. It consists of about 970 gm of uranium oxide and about 30 gm of fission products and transuranium actinides. Thus the initial ratio of:
(actinides + fission products) / (uranium oxide) = 30 / 970 = .0309

Using a separation process with a selection factor of Sf > 100 this material can be divided into Pile a and Pile b.

Let Na = number of gms of (actinides + fission products) in pile a.

Let Nb = number of gms of (actinides + fission products) in pile b.

Na + Nb = 30 gm

Let Ma = number of gms UO2 in pile a.

Let Mb = number of gms of UO2 in pile b

Ma + Mb = 970 gm

Assume reaction runs until:
Mb = 900 gm. Then:
Ma = 70 gm.

Assume a separation factor = Sf

Then (Na / Ma) = Sf (Nb / Mb) = Sf [(30 - Na) / (970 - Ma)]
[(970 - Ma) Na] = Sf [(30- Na) Ma]
(Sf - 1) (Ma Na) = 30 Sf Ma - 970 Na
(Sf - 1) = (30 Sf / Na) - (970 / Ma)
(30 Sf / Na) = (Sf - 1) + (970 / Ma)
Na = 30 Sf / [(Sf - 1) + (970 / Ma)]

(Na / Ma) = 30 Sf / [(Sf - 1) Ma + (970)]

For a separation factor of 100 we get:
(Na / Ma) = 30 Sf / [(Sf - 1) Ma + (970)] = 3000 / [99 (70) + 970]
= 3000 / 7900
= 0.3797

Thus the selective removal of 90% of the fuel mass in the form of uranium oxide achieved more than a 10 fold increase in actinide concentration.

Mb = 900 gm

Nb = 30 gm - Na
= 30 gm - {30 Sf / [(Sf - 1) + (970 / Ma)]}
= 30 - {3000 / [(99) + (970 / 70)]}
= 30 - {3000 / 112.8571}
= 3.417 gm

Nb / Mb = 3.417 / 900 = 0.003797

This result corresponds to about a 10 fold reduction in blanket rod material radio activity.

Clearly a separation factor of > 1000 would be much better in terms of reducing the radio activity of pile b.

Thus the material in Pile a, after oxide reduction and fission product extraction, is suitable for use in FNR core rods and the material in Pile b after oxide reduction is suitable for use in FNR blanket rods. The selectively removed uranium oxide, which constitutes over 90% of the blanket rod mass, and which will contain a small concentration of actinide impurity, is stored for future use as a fast neutron reactor blanket rod fuel material.

Subject to processing cost versus storage cost economics this separation process can be repeated to further lower the actinide concentration in the stored uranium oxide.

A breeder reactor fissions Pu-239 in its core to produce surplus neutrons, some of which migrate into and are absorbed by the blanket. Neutrons escaping the blanket are fully absorbed by the liquid sodium pool. In both the blanket and the core surplus neutrons convert U-238 into plutonium and other transuranium actinides. Periodically blanket and core rod reprocessing is used to discard fission products from the core rods and to transfer newly generated transuranium actinide atoms from the blanket rods to the core rods. The blanket rod mass is maintained by addition of new depleted UO2.

Zirconium, which tends to follow the fission products is extracted from the fission product stream and is recycled. The remaining extracted fission products, which generally have half lives of less than 30 years, are placed in isolated secure storage for 300 years to allow the radio toxicity of the fission products to naturally decay.

In a single multi-year fuel cycle a fast neutron reactor converts 10% to 20% of the initial core rod weight of a fuel bundle into fission products.

Each year a fractioon of the fuel bundles are reprocessed.

Uranium oxide is selectively removed from neutron activated blanket rod material to provide pure uranium oxide for forming new blanket rods. Part of the blanket rod material has converted into plutonium. Replacement UO2 for blanket rod formation is obtained from the U-238 inventory.

The transuranium actinides formed in the blanket rods are converted to metallic form and are sent to the core rod formation process.

Low atomic weight fission products and zirconium are extracted from the blanket rod residue and from the core rod material using a process known as pyro processing. Zirconium cladding introduced into the blanket rod process becomes approximately 20X more concentrated in the core rod process. The zirconium exits the pyroprocess along with the fission products. The zirconium is selectively extracted from the fission product stream using a dry chloride extraction process and the fission product residue is placed in engineered containers for 300+ year dry storage.

Zirconium is recycled through both the core rod and the blanket rod processes.

The high atomic weight material left behind after fission product extraction from core rods is a mixture of uranium and transuranium actinides. The extracted weight of fission products is replaced by an equal weight of new core rod material obtained from the blanket rod process. These materials are combined along with 10% zirconium to form new core rods. There is sufficient conversion of U-238 into new Pu-239 to maintain the Pu-239 concentration in the core rods required for reactor criticality.

Note that the plutonium is never chemically isolated.

The uranium isotope U-238 is naturally about 140 times more abundant than the uranium isotope U-235. However, U-235 is the primary fuel for water moderated nuclear reactors. As long as the rarer uranium isotope U-235 remains inexpensive the main financial justification for use of breeder reactors instead of water moderated reactors is safe disposal of the long lived and highly toxic transuranium actinides produced by water moderated nuclear reactors. However, the longer term justification for breeder reactors is to produce energy from U-238 without net production of transuranium actinides and without emission of CO2.

The FNR liquid sodium pool has a central reactor zone where the core fuel rods together form a critical mass and hence generate heat and emit neutrons. Also in the reactor zone are the blanket fuel rods surrounding the core rods which absorb excess neutrons emitted by the reactor core and convert U-238 into Pu-239 and other actinides. The material holding these rods in position is primarily HT-9 low nickel steel alloy tubing with B4C end caps.

The liquid sodium pool also has a perimeter zone where fuel bundles are stored to allow their radioactivity to decay to the point that thermal emission is not a problem in the subsequent fuel reprocessing steps.

In the Ottensmeyer Plan the spent CANDU fuel mass that is subject to energy intensive chemical processing is minimized by initial selective extraction of about 89.25% of the spent CANDU fuel mass weight in the form of UO2. This UO2 goes into storage for future use. About 81% of the residue is used to form initial blanket rods which are zirconium clad.

The remaining spent CANDU fuel material is processed to form FNR core rods. During this processing this material is converted to metallic form and is then separated into its high and low atomic weight components. Zirconium is then selectively extracted from the low atomic weight fission product stream. The fission products are sent to 300 year storage. The remaining high atomic weight uranium, plutonium, transuranic actinides and extracted zirconium are used in controlled ratios to form the fast neutron reactor (FNR) core rods.

The nominal initial FNR core rod composition by weight is:
70% U-238
20% Pu-239 and other actinides
10% zirconium

The nominal initial FNR blanket rod composition by weight is:
90% U-238
10% zirconium

The reactor core is where most of the fission reactions occur. The purpose of the blanket rods is to capture excess neutrons emitted by the reactor core and use these excess neutrons to breed more fissonable material for later use by the core. However, this new fissionable material can only move from the blanket rods to the core rods via periodic reprocessing of both the blanket rods and core the rods.

The first stage of the spent CANDU fuel separation process must be simple, nonenergy intensive and relatively inexpensive, because the first stage is used to process a large mass of material.

The fuel reprocessing must not introduce chemical impurites that inadvertently contaminate the process. This issue of possible chemical impurity contamination needs to be checked at an early date. The problem is that due to the large fuel mass of over 1,000 tonnes that must be reduced to metallic form it may be uneconomic to use reagent quality aluminum or magnesium for chemical reduction of high atomic weight oxides to metals. Industrial quality reduction metals must be carefully checked to ensure that they do not introduce chemical impurities that adversely affect the overall process. For example if the reduction metal contains calcium and if that calcium leaks into the metallic fuel it will form Ca-41 which is a long lived isotope that is difficult and expensive to dispose of safely.

In the Ottensmeyer Plan the spent CANDU fuel or recycled blanket rod material is first chopped up and then dissolved in nitric acid (HNO3). The solution is cooled and about 89.25% of the solute weight forms urynal nitrate hexahydrate (UO2(NO3)2-6H2O) crystals that are then physically extracted from the solution. Reheating these crystals liberates nitric acid vapor which is recondensed for nitric acid recovery. Left behind is nearly pure uranium oxide (a mixture of UO2 and U3O8). This extracted uranium oxide still contains a small concentration of fission products and actinides but this extracted uranium oxide is suitable for future use for forming FNR blanket rods. The extacted uranium oxide should be stored in engineered containers for future use as blanket rod material.

The nitric acid process is repeated on the residue to extract about 81% of the residue weight as UO2 which is reduced to uranium metal and then converted into blanket rods. The blanket rods are initially 90% U + 10% Zr. The blanket rods initally contain no fissionable material.

The first step in forming FNR fuel is to use a nitric acid process to separate UO2 from spent CANDU fuel, as set out above. The spent CANDU fuel residue is an input to the core rod and blanket rod fabrication processes.

The core rods initially contain about 20% by weight of fissionable material, principally plutonium. The balance is about 70% U-238 and 10% zirconium.

The fast neutron reactor is operated continuously except for brief annual shutdowns to permit exchange and repositioning of fuel bundles.

During one fuel cycle, which is typically 25 years in duration, the reactor converts about 20% of the core rod weight into fission products and converts about 7% of the blanket rod weight into fissionable material.

After one fuel cycle each fuel bundle is stored for part of another fuel cycle outside the neutron flux to allow the bundle's fission product radioactivity to decay. The fuel bundle is then removed for fuel rod reprocessing. Let N be the number of years per fuel cycle. The fraction (1 / N) of the fuel bundles is reprocessed every year so that the reprocessing operation is nearly continuous.

Material recycling does not commence until a FNR has operated for about N years and does not reach its maximum level in terms of processed material radioactivity until after about 2 N years of FNR operation.

The following steps are done by remote manipulation and/or robotics in an argon atmosphere.

Material recycling starts with removal of a fuel bundle from the perimeter zone of the liquid sodium pool. The fuel bundle is lifted out of the pool and allowed to drip dry.

By this time the fuel bundle thermal emission has decreased enough that when the bundle is removed from the liquid sodium pool its heat emission can be readily removed with atmospheric pressure naturally circulating argon gas.

The fuel bundle is then moved to a room temperature space where in a argon atmosphere it is disassembled and sorted into its major component parts which are:
blanket rods,
core rods,
sodium salts,
inert gases

Each of these components is processes in a different manner.

The steel tube material will have cumulative fast neutron damage and must be shipped to a specialty steel tube manufacturer for recycling. There may be an issue of gradual accumulation of radio daughters in the steel that must be chemically removed to enable material recycling.

The blanket rods are moved to the blanket rod pre-processing facility which extracts much of the uranium oxide.

After blanket rod pre-processing the remaining blanket rod mass residue is transported to a remote site such as Chalk River for further processing and new fuel rod fabrication.

The liquid sodium is saved for reuse in future fuel tubes.

After one fuel cycle time the core and blanket rods are removed from the FNR and are pre-processed. The blanket rods, which in a FNR produce Pu-239 and other actinides, are reprocessed to provide both pure UO2 for replacement blanket rods and concentrated Pu-239 plus other actinides plus zirconium for use in replacement core rods.

Replacement blanket rods are produced on the FNR site. The material containing concentrated Pu-239 plus other actinides plus zirconium is shipped to a remote site such as Chalk River where it is used to produce new metallic core rods.

The blanket rods are about 75% of the total FNR fuel weight. A newly formed blanket rod has a relatively low radioactivity. Blanket fuel rods can never form a critical mass and hence can be safely processed and interim stored on a reactor site to minimize material transportation costs.

FNR core fuel rods, if mishandled by a negligent or malevolent party, could potentially form a critical mass. Hence, for public safety the irradiated core rods containing U, Pu-239 plus other transuranium actinides plus zirconium plus fission products are shipped to a remote reprocessing site such as Chalk River, that is far from any major urban center. At that site the fission products are extracted and the new fissionable material produced in the blanket rods is used to form new enriched metallic core rods.

The low atomic weight fission products plus zirconium are extracted from the FNR core rods using a highly selective process known as pyro processing.

Blanket rod processing is used to selectively extract part of the U and part of the Zr from the blanket rod material to convert the residue into core rod feedstock. The extracted material is used to form new blanket rods.

The residue from the blanket rod process is sent to the core rod process which is conducted at a remote site. In the core rod process this material is reduced to metallic form and combined with recycled core rod material. A pyro process is used to separate fission products and zirconium from the high atomic weight atoms. Then a dry chlorination process is used to separate the zirconium from the fission products for zirconium recycling.

The high atomic weight atoms are combined with an appropriate amount of recycled zirconium to form new core fuel rods. There is some over production of core rod material which is sent to inventory for future use as start fuel for another FNR. There is some left over zirconium that is sent to inventory for future use in the blanket rods.

The fission products are sent to 300 year storage.

In the fuel rod recycling process there is about a 7% loss of blanket rod weight per fuel cycle a which is replaced using U-238 drawn from inventory. Note that the blanket rod weight is about 4X the core rod weight, so a loss of 7% of the blanket rod weight is an input to the core rod process equal to about 21% ______of the core rod weight. This input offsets the loss of core rod weight due to extraction of fission products and due to over production of core rod material.

An atomic weight based process with a very high separation ratio, known as pyroprocessing, is used to separate a unit weight of fuel rod material into high atomic weight and low atomic weight fractions. This process can be much more elaborate than the blanket rod process step because this process handles only a small fraction as much material.

The primary atomic weight separation mechanism is liquid cadmium flotation/distillation. With a few exceptions atoms with atomic weights less than cadmium (atomic weight = 112) have half lives less than 30 years and hence are are short lived fission products or zirconium. The remaining high atomic weight atoms are recycled into the core rods.

Then zirconium is extracted from the fission products using a dry chlorination process that forms ZrCl4 + a range of other chlorides. These chlorides are distilled to concentrate the ZrCl4 and then sodium is used to reduce the ZrCl4 to Zr which allows recovery of most of the zirconium. The resulting NaCl is electrolytically separated back into Na and Cl.

The uranium, zirconium and actinides are recycled in controlled ratios and formed into as FNR core rods. Part of the zirconium is shipped back to the reactor site to be incorporated into future blanket rods.

About 90% of the new core rods are shipped back to the reactor site. The remaining 10% of the core rods produced are surplus to current requirements and go into storage pending their use as start fuel for another FNR. These extra core rods produced at the remote processing facility are placed in engineered containers packed with B4C and are shipped to a naturally dry accessible depleted hardrock mine for secure isolated storage.

The dry chlorinated fission products, which mostly have half lives of less than 30 years, together with a small amount of residual zirconium, are placed in secure isolated storage for about 300 years to allow the fission products to naturally decay into stable isotopes. The neutron activated zirconium isotopes decay in about 3 months through niobium isotopes to become stable molybdenum isotopes. Note that the chlorinated fission products are potentially water soluble and must be completely excluded from ground water for at least 300 years.

A very small fraction of the low atomic weight fission product atoms are long lived isotopes such as Sn-126 (half-life ~ 100,000 years). After a container of fission products has been stored for 300 years the radio activity of the container is dominated by a few long lived low atomic weight isotopes. The container is scanned with a gamma ray spectrometer. Based on the recorded gamma ray spectrum chemical treatments are used to selectively separate the long lived isotopes from the decayed fission products. These extracted long lived isotopes can be either subjected to further fast neutron irradiation or can be sent to very long term storage.

Note that any material to be reirradiated by neutrons must first be completely dechlorinated to prevent conversion of Cl-35 into long lived Cl-36.

A second neutron irradiation stage of processing is effective at suppressing the long lived fission products Tc-99, Sn-126, Se-79, Zr-93, Cs-135, Pd-107 and I-129. After a second 300 year storage cycle the container contents are again scanned with a gamma ray spectrometer and again any remaining long lived isotopes can be selectively extracted.

When the radio activity of a container of processed waste is less than the radio activity of an identical container of natural uranium, the container contents can be safely removed and recycled. The fission product spectrum contains elements such as the rare earths that have high commercial value. The released DGR storage space and the released storage containers are then available for reuse.

Each fuel bundle is clearly and uniquely numbered and is clearly labelled as to whether it is an active bundle containing core rods or a passive bundle containing just blanket rods.

If there were no fuel bundle changes, over about a ten year period fission products gradually accumulate in the core rods of the reactor to the point that reactor criticality can no longer be maintained. Hence a program of ongoing fuel bundle changes is used to maintain reactor criticality while minimizing the total required fuel inventory and while controlling the total fast neutron exposure of each fuel bundle.

This program of fuel bundle repositioning is essential to minimize the cost of the FNR fuel inventory and fuel processing. However, it does leave open the possibility that an unsupervised persons with an FNR could use some of the blanket fuel bundles for breeding military grade plutonium. Hence the movements of each fuel bundle should be automatically tracked. An alarm should be initiated on any violation of the first-in first-out fuel bundle exposure sequence.

About once a year the FNR is briefly shut down by withdrawal of every second active fuel bundle. Then using the gantry crane one at a time each selected fuel bundle is lifted vertically sufficiently for its lower support rod to clear the fuel bundle support frame and then, while the fuel bundle is still immersed in liquid sodium, The fuel bundle is moved horizontally to a new fuel bundle position near the perimeter of the liquid sodium pool. The fuel bundle is reinserted into the support frame at that position. The square float over the fuel bundle travels with the fuel bundle.

Note that to access active fuel bundles it is necessary to temporarily move a few passive fuel bundles out of the way.

By a suitable sequence of fuel bundle moves any desired repositioning of the fuel bundles can be achieved. However, it is important to keep a complete record of which fuel bundle was where and for how long, so that the total neutron exposure of each fuel bundle can easily be calculated and the first in first out exposure sequence maintained.

Let N be the number of years in a fuel cycle. The fuelling concept is that every year about 1 / N of the fuel bundles are moved from the reactor zone to the perimeter zone of the sodium pool. The empty fuel bundle positions in the reactor zone are filled with new fuel bundles. Every year the fraction (1 / N) of the fuel bundles in the perimeter zone is removed for recycling creating space in the perimeter zone to allow continuing movement of fuel bundles from the reactor zone into the perimeter zone. The fuel bundles are removed from the perimeter zone on a first-in first-out basis, so that each fuel bundle remains in that zone for several years.

Meanwhile the fuel bundles that are removed for recycling are reprocessed into new fuel bundles for use the following year. Hence commencing about N years after reactor startup there is a continuous material flow through the recycling process. The recycling process must be sized to handle at least (1 / N) of the standard FNR fuel load every year. For every FNR in service the core fuel rod process capacity must be proportionately scaled up.

Each FNR requires a partial extra fuel load so that at any instant in time one complete fuel load is in the reactor zone supplying energy and the partial fuel load fuel bundles are either in the perimeter zone cooling or are in the recycling process.

Note that the total required fuel inventory is nearly constant independent of the material flow rate through the recycling process. For example, if 20% of the fuel is recycled each year there is an ongoing requirement for at least 1.2 complete fuel loads for each FNR.

Note that the flow of fuel material through the reprocessing system is nearly constant and is proportional to reactor power but is otherwise independent of the details of the reactor design.

The required start fuel inventory per reactor is a major constraint on the rate at which FNRs can be deployed. It appears that at this time there is only enough spent CANDU fuel to support 4 CANDU 6E size FNRS. However, since the CANDU fleet will operate for many years into the future, by the time the FNR fleet exceeds 4 CANDU 6E size reactors, there will probably be enough additional spent CANDU fuel to start at least another 4 FNRs.

This fuel reprocessing cycle is repeated over and over for both blanket rods and core rods until all of the available high atomic weight atoms contained in the fuel inventory are consumed.

The FNR fuel recycling process increases the fraction of CANDU fuel that fissions from less than 1% to almost 100%. This process yields sufficient energy from the existing CANDU spent fuel to energize the projected Canadian FNR fleet for hundreds of years.

While in 300 year storage the fission products naturally decay to become primarily stable elements with high economic value. The fission product stream contains a few low atomic weight elements with isotopes that have long half lives. These isotopes may have to be selectively chemically extracted after 300 years in storage. Such selectively extracted radioactive isotopes could either be reprocessed or could be placed in long term DGR storage.

Note that in the aforementioned process there are ongoing requirements for interim storage of substantial inventories of highly radioactive materials including:
fission products;
spent CANDU fuel;
new core fuel rods;
used core fuel rods;
new blanket fuel rods;
used blanket fuel rods;
neutron activated zirconium;
neutron activated steel;
neutron activated B4C

These higly radioactive materials, if not immediately required, should be stored in engineered containers located in accessible secure safe dry storage such as a naturally dry depleted hardrock mine.

Double wall sealed storage containers (outer wall porcelain, inner wall stainless steel, dielectric between the inner and outer walls) should be used to store the radioactive material while it is in the DGR. While in the DGR the physical integrity of each nuclear waste storage container is constantly remotely monitored by sensing the container temperature, dielectric level and dielectric loss tangent. Any container wall failure is detected via a dielectric level change long before there is any radioactive material leakage.

The accessible naturally dry gravity drained DGR should have sumps such that if there is a radio isotope leakage it is detected at a sump and remedial measures can be implemented long before there is any leak of radio isotopes into the external environment. Since the Ottensmeyer Plan only requires a few hundred years for full implementation, the containers do not have to be designed to last for more than a few thousand years.

The accessible naturally dry (gravity drained) Deep Geologic Repository (DGR) should be at least 300 m above the local water table and should be naturally ventilated. Resistance to DGR damage by glaciation and other events is ensured by forming the DGR inside the high density granite core of a mountain that provides over 400 m of overhead rock above the DGR. There should be no attempt to find a "receptive community" because at any geologically suitable DGR site there will be insufficient fresh water to support a permanent community. The DGR location should be determined primarily by favorable geophysical criteria, not by political criteria.

The mine complex known as Jersey Emerald, in British Columbia, meets this requirement and may be available at a modest price.

There may be other less favorable but still acceptable storage site alternatives elsewhere in Canada such as in depleted uranium mines in the vicinity of Elliot Lake, Ontario.

This web page last updated June 11, 2017

Home Energy Nuclear Electricity Climate Change Lighting Control Contacts Links