Visit to Marviken

Sometimes the NPYP bunch tires of being rough on the roughnecks and decides to leave the jungle headquarters for a walk on the streets like normal people. Such an momentous occasion happened this week, and where else could they be found other than at the only oil fired nuclear power plant in the world (don't tell us we don't know how to party!)?

The power plant in question is Marviken, located about 150 km driving distance from Stockholm. It was supposed to be the first large scale electricity producing nuclear power plant built according to "the Swedish line". The design principles behind the Swedish line was:

  • Natural uranium as fuel so that the abundant Swedish uranium resources could be utilized without any need to depend on import.
  • Heavy water as moderator because light water steals to many neutrons to be possible to use with natural uranium fuel.
  • Possibility to refuel during operation so that fuel bundles can be removed at the point where the Plutonium isotope composition is the most beneficial as weapons material.

Three reactors where constructed based on the Swedish line. The construction of the first reactor, R1, was started in 1951 and it was placed in a dug out cavern below Kungliga Tekniska Högskolan (the Royal Institute of Technology in Stockholm). It was a small experimental reactor with a thermal power of 1 MW. The purpose of R1 was not to produce power or plutonium but to gain insight into reactor physics. The second reactor belonging to the Swedish line was R3 (R2 was a light water material test reactor), or Ågesta, a smaller reactor designed to deliver 55 MW of district heating and 10 MW of electricity.

The third reactor, Marviken, went through many changes before the final design. At first it was designed to be a 100 MW pressurized heavy water reactor, with a secondary circuit where steam is produced from normal water. The fuel shuffling machine was supposed to be within the reactor vessel and the goal was to be able to shuffle and replace fuel during lower power operation. But parallel to the pressurized designed work was ongoing on a larger design.

The reactor tank within the containment

The new design brought many significant changes, the larger design was supposed to run on slightly enriched uranium (even though lower power operation on natural uranium also was possible) and most significantly it was a boiling heavy water reactor. The heavy water was supposed to be boiled within the reactor vessel and then directly go to the turbine like in a BWR. At the same time ideas were discussed about superheating the steam within the vessel. The idea behind superheating is that the water flows in through the bottom of the vessel, gets turned into steam while passing up through the fuel assemblies. Then at the top of the reactor the steam flow is reverted and the steam goes back down through a bunch of special fuel assemblies where the steam is heated up to 500 degrees Celsius.

Picture from SKI 2007-18

The problem was that no one in the world had built a boiling heavy water reactor and superheating was a new concept not yet tested. The advantage with superheating is that it increases the efficiency of the plant and helps with turbine design because the steam temperature would be similar to the temperature in fossil power plants. The drawback is that superheating places large demands on the materials in the fuel assemblies and the direct cycle meant that a turbine had to be developed to run on heavy water steam. A further consideration for the turbine was that water leakage had to be kept to an absolute minimum due to the high cost of heavy water. The combination of boiling heavy water, internal reshuffling machine and superheating lead to a very complicated design that raised alarms already from the beginning. The optimists won the discussions though and in 1965 the construction of the plant started.

During the construction the operator of the plant (Vattenfall) got more and more doubtful about the project. Vattenfall started attacking both the ideas of superheating and the reshuffling machine. All the doubt and arguments lead to a point where Vattenfall refused any responsibility for the project. The engineers at Ågesta also raised serious doubt about Marviken and wrote a report stating that Marviken is a project that is impossible to realize.


Random gauges in the control room

However construction proceded. ASEA and AB Atomenergi (the two companies responsible for construction of the plant, the second one a wholly state owned company) wanted to proceed with the superheating as a competitive advantage against the emerging American light water reactors. In 1968 the plant had reached a point where hot tests with normal water could be done, but another problem arose; during the end of the 1960's the American Atomic Energy Commission released standards for reactor safety that quickly became world standards. In Sweden it was realised that Marviken would not live up to those standards. To make matters even worse, new and more detailed calculations showed that the plant would have slightly positive reactivity coefficients *. This was in contrast to earlier analysis that showed they would be negative.


The turbine

By 1969 the plans for superheating was abandoned. The technical challenges had not been overcome and other superheating projects around the world were also scrapped. During the hot test runs other problems appeared and AB Atomenergi finally suggested that the project should be cancelled. The government went along with the suggestions, and the heavy water and already manufactured fuel assemblies were sold. At the end of the day 640 million Swedish crowns has been spent on the project. To recover some of the losses another building was erected where a 200 MW oil boiler was installed. The boiler could use the turbine and other systems already built for Marviken.

In a weird twist of faith this turned out to be the start of the real interesting stuff. Marviken was built with the same type of wet containment as BWR's. The principles behind the wet containment is that if there is a sudden break in the reactor circuit the steam will blow out into the dry part of the containment, then through pipes it will be directed into a huge pool of water (called a pressure supression pool). When the steam goes through the water it condenses and the pressure of the containment system remains manageable. This was a golden opportunity to test the containment, nowhere else in the world was there a full scale containment available for experiments.


Inside the drywell, the bottom of the reactor vessel can be hinted at the top of the picture

Not only Sweden was interested, companies from all over the world (primarily US, Japan, and Germany) came to Marviken to collaborate on several large scale tests. The pressure suppression pools were tested to maximum capacity, containment venting was tested in a spectacular experiment where over 20 tons of steam per second was blown straight out of the reactor. Materials were melted in the reactor to test how materials from a meltdown would act within the containment. Very important data was also gathered on things like critical flow (the maximum flow that can go through a given pipe break size), and oscillations in the pressure suppression system. All kinds of theoretical models were verified and shown to be accurate enough to be used in safety analysis of other plants, not to mention the experience gained with how different gauges and components can handle such events.


On top of the fuel pool

Visiting this marvelous place was amazing, our guide to the plant had worked there since just after it was decided the plant will be converted to a oil boiler. One could not help but to be impressed by his incredible knowledge of the plant. Few people now a days have that kind of feel for the facility, imagine a job where you sit in the control room at one moment and the next you are out in the plant replacing a broken bearing or walking around manually adjusting valves. Just knowing which valve you need to adjust is an art by itself.

Kudos to an interesting man in an interesting place!

For more information about Marviken (unfortunately in Swedish only) see this SKI report: Marvikenreaktorn - ett industripolitiskt utvecklingsprojekt i otakt med tiden and this article Marviken sätter punkt.

* Reactivity feedbacks are natural feedback loops that occur when the temperature of materials in a reactor changes. For instance if the fuel temperature increases some fuel isotopes can absorb more neutrons, resulting in less neutrons being available for fission and thus the power goes down. There are several contributions to the total reactivity feedback and it should always be negative so that the reactor is self controlling.


Tags: , , , , , , ,