ML-1 Mobile Power System: Reactor in a Box

At Adams Atomic Engines, Inc. we like the concept of the closed cycle gas turbine enough to be developing an updated design that builds on the lessons learned from past experiences.
The ML-1 experimental reactor was unique. It was not a pressurized water reactor with a steam energy conversion system. Instead, ML-1 was the first nitrogen cooled, water moderated reactor with a nitrogen turbine energy conversion system. Its major design criteria was compactness.

ML-1 could be packed into 6 shipping containers – one for the reactor, one for the complete heat conversion system, one for the control room, and three others for cabling, auxiliary gas storage and handling equipment and miscellaneous tools and critical supplies.

The two major containers weighed 15 tons each while the four additional containers each weighed between three and four tons. They were designed to fit into any of the Army’s transport systems including C-130 aircraft, standard Army trucks, and rail.

In order to reduce the weight of shielding needing transport, the reactor was designed to be installed with a human exclusion boundary of 500 feet.

Design Challenges

In order to minimize the engine volume and mass, the decision was made to operate the engine with nitrogen pressurized to approximately 9 bar – 9 times normal atmospheric pressure – at the compressor inlet. This decision, though it helped reduce the size of the heat exchangers and turbomachinery somewhat, made the design uniquely difficult.

Essentially every other gas turbine ever built has operated with air at atmospheric pressure as the working fluid. The designers of ML-1 had the difficult challenge of making the machine perform as desired with a high density working fluid. This requires the reduction of critical machine clearances and makes accurate balancing far more critical for long term, reliable operation.

A second design decision that made the engine construction more challenging than required was the decision to add a recuperator to the system. Though recuperators have proven that they can improve gas turbine efficiency by several percent in stationary applications, they are not normally used in mobile engines because the additional heat exchanger adds more weight and space than it is worth.

The reactor heat system also required a stretch of existing technology. In order to minimize the size of the reactor, designers decided to use water inside pressure tubes as the neutron moderator. In order to prevent boiling, the water in the tubes was circulated to maintain the temperature below 250 F.

The water tubes were interspersed throughout the core between the fuel bundles. The nitrogen gas flowed past both the water tubes and the fuel bundles and ranged in temperature from 800 F at the core inlet to 1200 F at the outlet. The physical distance between the inlet and the outlet was less than two feet; the temperature extremes made material selection very important.

Testing Experience

The designers of the ML-1 decided to test two different heat engines that could each be connected to the reactor heat source. Once of the machines had an 11 stage axial flow compressor designed and constructed by Fairchild-Stratos Corporation while the other included a two stage centrifugal flow compressor designed and built by Clark Brothers Company.

Neither heat engine was able to meet its designed power output because neither compressor was able to produce the required flow at the required differential pressure. Rather than achieving a power output of 300 kw the best that the tested system could achieve was less than 200 kw. Engineering evaluations were made indicating that some minor adjustments could be made that would raise the performance of the machine, but it is not apparent from the historical record that this kind of rework was ever completed.

A second problem that surfaced during the testing program was related to the moderator water tubes. The high thermal and temperature stress of the tubes combined with manufacturing flaws to cause cracking in the tube welds. The cracks allowed water to enter into the coolant system and required a lengthy hiatus in the test program to correct the problem.

A final problem that had a major effect on the system was the failure of the internal insulation of the regenerator. This was installed under the assumption that it would reduce heat losses and thus improve performance. The insulation consisted of a blanket of fine particles covered with a metal foil. The foil tore loose because of aerodynamic buffeting during testing, causing the distribution of the fine particles throughout the system. After the dust was removed from the engine, testing continued without the insulation.

Lessons Learned

Though the difficulties experienced by the ML-1 testers were the type that are common with the first of a kind of any complex piece of machinery, they proved to be fatal for the program and helped destroy any budding interest in nuclear gas turbines.

Because of the increasing amount of money needed to fight the Vietnam War, the Army’s research and development budget for non weapons items was severely constrained. There was little support for funding experimental nuclear systems in 1963, particularly experimental systems that seemed to have so many difficulties that needed fixing.

Now, however, after more than 30 years of technological developments, it is worth summarizing the lessons that can be learned for future closed cycle gas turbine development.

  • Pressurizing gas turbine cycles may be a good idea on paper, but there are practical engineering difficulties that must be overcome if it is to be used in a real system.
  • Recuperators are troublesome, particularly if space and weight are constraining factors in system design.
  • Water tube reactors are unnecessarily complex, particularly since there have been excellent results achieved by high temperature graphite moderated reactors.
  • Any material that can potentially contaminate a closed cycle turbine system should be avoided.
  • If possible, well-proven components should be integrated into a complete system rather than designing each component from scratch.

About Rod Adams