Search This Blog

Sunday, July 13, 2014

CANDU reactors

I had intended to write up a post about NMRI, or Nuclear Magnetic Resonance Imaging.  I may or may not get to that post.  To be frank, I don't find flipping atoms back and forth in a big magnet quite as interesting as discovering X-rays or splitting atoms.

So instead of a post on NMRI, I am going to write a post about splitting atoms :)

One reactor type in use today that I didn't mention earlier is called the CANDU reactor.  It's an acronym for CANada Deuterium Uranium reactor.

This unique Canadian reactor design came about for several reasons: 
  1. At the end of World War 2, the entire world knew about nuclear weapons and was aglow with the possibility of using the atom for peaceful uses like cheap electrical power - yeah right!
  2. Canada had no facility for enriching Uranium, nor the financial resources to build one.
  3. Neither did Canada have the manufacturing facilities to build large, high quality pressure vessels, such as those housing PWR and BWR reactor cores.
As a result of the above issues, the Canadians came up with a reactor design that uses heavy water for a neutron moderator - water in which each Hydrogen atom carries an extra neutron.  Heavy water moderation allows the use of unenriched Uranium fuel.

Recall that neutrons produced from fission have very high energy levels (speed), and must be moderated (slowed down) by collisions with other atoms.  With ordinary water there is a low, but significant probability that neutrons will be captured by the hydrogen atom (which becomes a deuterium atom), and be lost to the fission process.  This slight amount of neutron capture by normal hydrogen atoms prevents the use of normal water in an unenriched Uranium reactor.

However Heavy water hydrogen atoms already have a neutron, and the probability that a deuterium atom will capture a neutron to become Tritium, is very, very low.  Thus in the core of a heavy water reactor, enough neutrons are available to cause fission in unenriched Uranium.

Below, the three isotopes of hydrogen - courtesy Dirk Hünniger, German Wiki.

It is considerably easier to separate normal deuterium from ordinary hydrogen than it is to separate U-235 from U-238 (See the earlier post on Uranium Enrichment).  The far simpler separation of heavy water from normal water allows the creation of a reactor that uses naturally occurring uranium fuel. 

The CANDU design is similar to the graphite reactors that were used for Plutonium production, including horizontal fuel assemblies and the provision for online refueling.  This has the huge economic advantage of increasing reactor availability.  

To refuel a PWR or BWR, the reactor must be shut down, and cooled down.  The reactor vessel head is removed, the inner fuel bundles are painstakingly removed, the outer fuel bundles are moved to the center, and a 1/3 load of fresh fuel is installed at the perimiter.  The primary coolant system is closed up, and a lengthy re-start process takes place.  This tedious refueling process typically takes the reactor offline for 4 weeks approximately every 18 months, reducing availability.

Below, a PWR undergoing refueling.  A spent fuel element (under several feet of water) glowing with Cerenkov Radiation is being moved to the spent fuel pool.

Back to the very cool CANDU design...  Let's take a look.  This is a very clever design because it eliminates several of the issues that PWR and BWR reactors have. 

CANDU cutaway, courtesy Wiki:

  1. Fuel, Ceramic Uranium Oxide, clad in Zircaloy.
  2. Calandria.  A low-pressure vessel filled with heavy water, with penetrations for high pressure fuel channels and control rods.  Is filled with heavy water moderator.
  3. Control rods to start and stop the nuclear reaction.
  4. Pressurizer vessel.  Prevents in-core boiling by raising pressure above the boiling point of temperatures in the reactor.  Also provides a surge volume for coolant temperature changes.
  5. Steam generator.  Uses heat from the radioactive primary coolant to boil non-radioactive water for a Rankine cycle power plant.
  6. Feedwater pump.  Pumps normal water into the steam generator to be boiled.
  7. Primary coolant pump.  Pumps Heavy water to the steam generator and back to the reactor to pick up more heat.
  8. Online refueling apparatus
  9. Heavy water moderator
  10. Pressure tube that holds fuel and through which high pressure heavy water primary coolant circulates
  11. Steam from steam generator (non radioactive)
  12. Feedwater to steam generator (non radioactive)
  13. Containment structure 
Below, fuel bundles for a CANDU reactor. Originally uploaded to English Wikipedia by Dr. Jeremy Whitlock of AECL

Below, a view of the calandria, which holds heavy water, and through which the pressure tubes (containing fuel bundles) pass.  This reactor is undergoing a major repair or refit.  All the risers leading from the pressure tubes to the steam generators have been cut off and capped for the duration of the repair.

The CANDU reactors have a couple of undesirable characteristics from a reactor safety standpoint, and those are a positive void coefficient of reactivity, and a positive power coefficient of reactivity.  Both are positive feedback loops that can accelerate a runaway reaction. 

The Wiki article on CANDU reactors is overwhelmingly positive, making me wonder (too lazy to look it up) how much of it is simply marketing content from the AECL (Atomic Energy of Canada, Limited).  A great deal of the article's content is about sales, of all things. 

There have been two nasty accidents at CANDU type reactors, and neither of these is mentioned in the Wiki article about them, although I will grant that both the accidents occurred on research reactors rather than on those used by utilities to make electrical power.

The first accident was a power excursion that ruptured the primary coolant loop.  It occurred at a small research reactor called NRX.  This was a 10 MW (thermal output) reactor that was eventually upgraded to 42 MW.  There is a Wikipedia article about NRX, but oddly enough, you have to know about the accident in advance to look it up, because there are no links to it on the CANDU wiki :)
Edit :  Ah, I see the Wiki has been changed, with links to the NRX and NRU articles added on December 3, 2014.  No doubt this was a result of my highly influential blog ;) hahahaha!

Here's the relevant stuff from the Wiki article about NRX:

On December 12, 1952, the NRX reactor suffered a partial meltdown due to operator error and mechanical problems in the shut-off systems. For test purposes, some of the tubes were disconnected from high pressure water cooling and connected by hoses to a temporary cooling system and one was cooled only by airflow.
During tests on low power, with low coolant flux through the core, the supervisor noticed several control rods being pulled from the core, and found an operator in the basement opening pneumatic valves. Wrongly opened valves were immediately closed, but some of the control rods did not reenter the core and stuck in almost withdrawn positions, but still low enough for their status lights to indicate them as lowered.
Due to a miscommunication between the supervisor and the control room operator, wrong buttons were pressed when the supervisor asked for lowering the control rods into the core. Instead of sealing the withdrawn control rods to the pneumatic system, the safeguard bank of four control rods was withdrawn from the core. Operator however noticed that the power level exponentially increases, doubling each 2 seconds, and tripped the reactor. Three of the safeguard control rods however were not inserted into the core and the fourth took abnormally long time, about 90 seconds, to slide back, while the power kept rising. After just 10 seconds 17 MW were reached.
The cooling water boiled in the tubes connected to the temporary cooling system, and some of them ruptured; the positive void coefficient of the reactor led to yet higher power increase rate. About 14 seconds later valves were opened to drain the heavy water from the calandria, which however took some time; power increased for 5 more seconds, peaked at 80 MW, then went down as the moderator level decreased and was at zero 25 seconds later.
Meanwhile some fuel elements melted and the calandria was pierced at several places; helium leaked and air was aspired inside. Hydrogen and other gases evolved by high-temperature reaction of metals with cooling water, and 3–4 minutes later oxyhydrogen exploded in the calandria.
During the incident, some gaseous fission products were vented to the atmosphere and heavy water in calandria was contaminated with the cooling water and the fission products. 
To remove decay heat, the water cooling system was kept operating, leaking contaminated coolant to the floor. About 10 kilocuries (400 TBq) of radioactive materials, contained in about a million gallons (about 4000 m3) of water, were dumped to the basement of the reactor building during next few days.
Clean-up of the site required several months of work, partially carried out by 150 US Navy personnel who had been training in the area, including future US president Jimmy Carter.[7] The NRX reactor core and calandria, damaged beyond repair, were removed and buried, and an improved replacement was installed; the refurbished reactor was operating again within two years.[8]
The lessons learned in the 1952 accident advanced the field of reactor safety significantly, and the concepts it highlighted (diversity and independence of safety systems, guaranteed shutdown capability, efficiency of man-machine interface) became fundamentals of reactor design.

The second reactor accident involving a CANDU-type reactor involved the NRU reactor.  It was initially designed as a 200 MW reactor, fueled with natural Uranium.  Here is the Wiki article for that accident.
The second accident, in 1958, involved a fuel rupture and fire in the NRU reactor building. Some fuel-rods were overheated. With a robotic crane, one of the rods with metallic uranium was pulled out of the reactor vessel. When the arm of the crane moved away from the vessel, the uranium caught fire and the rod broke. The largest part of the rod fell down into the containment vessel, still burning. The whole building was contaminated.
The valves of the ventilation-system were opened and a large area outside the building was contaminated. The fire was extinguished by scientists and maintenance-men in protective clothing running along the hole in the containment vessel with buckets of wet sand, throwing the sand down at the moment they passed the smoking entrance.[7]
Both accidents required a major cleanup effort involving many civilian and military personnel. Follow-up health monitoring of these workers has not revealed any adverse impacts from the two accidents.[8][9] However, the Canadian Coalition for Nuclear Responsibility, an anti-nuclear watchdog group, notes that some cleanup workers who were part of the military contingent assigned to the NRU reactor building unsuccessfully applied for a military disability pension due to health damages.[4]
And that, as they say, is the "Other side of the story"!  However I have to say these reactors have a fantastic safety record, compared to pretty much any other reactor design out there.  That view still has to be tempered with how few of these are in operation when compared with the others.

Eight CANDU reactors provide the steam for the world's second-largest power plant by nameplate capacity.  That is the 6232 Megawatt Bruce Station, on the North Shore of Lake Huron.
Below is a picture of the turbine hall of Bruce A.  Impressive...

No comments: