Designing A Maintainable Tokamak Reactor *

L. M. Waganer, F. R. Cole, and the ARIES Team **

McDonnell Douglas Aerospace, McDonnell Douglas Corporation
P.O. Box 516, St. Louis, MO 63166-0516, USA

Presented at the 19th Symposium on Fusion Technology
Lisbon, Portugal
16-20 September 1996

The long-standing goal for controlled nuclear fusion has been the generation of affordable electric power. There are other significant benefits of fusion, such as unlimited fuel, lower levels of radioactive waste products, and eased restrictions on licensing of new plants. But the economics of constructing and operating a fusion power plant has been a major hurdle to overcome. Designing and building a maintainable tokamak reactor is a key element in the ultimate feasibility of the fusion power plant. This integrated fusion power plant design will significantly improve the maintenance of the plant.


The U.S. ARIES Team [1] is developing a design approach to achieve high levels of maintainability and availability. Tokamak reactors have shown the best promise of delivering a controlled fusion plasma suitable for power production. However, placement of toroidal and poloidal field coils restricts the access to, and removal of, the life-limited power core components. Modified coil geometries and construction techniques provide improved maintenance access.

The entire power core and supporting systems must be designed from the beginning to achieve high levels of maintainability. In ARIES-RS [2], large modules of the power core are removable with a minimum number of disconnects. Large openings in the vacuum vessel allow the free movement of the removed components. Features are included to easily and quickly open the cryostat and vacuum vessel, disconnect the structural attachments, and guide the removal of the components. Maintenance equipment capable of carefully, but quickly, handling large components will be required. The maintenance will likely be largely autonomous with minimal human intervention.


There is a strong incentive to remove the first wall, blanket, and divertor components quickly in order to place the power core back into operation as soon as possible. The goal is to remove a large sector of the power core that contains those components that need to be replaced most frequently. The fewest number of toroidal field (TF) coils would offer the potential for the largest sector to be removed, but toroidal field ripple considerations set a practical limit of 16 coils.

Space required to remove a 1/16-sized sector of the first wall, blanket, and divertor (FWBD) forces placement of the outboard TF coil legs farther from the plasma. Under constant tension, this results in a larger TF coil vertical size and, consequently, very large PF coils. The entire coil system becomes unmanageable.

Each 1/16th sector could be broken into two or more subsectors, but the complexity of the handling, plumbing, and shielding reduces the attractiveness of this option. Instead, the project decided to use full-sector replacement modules and modify the shape of the TF coils by flattening the top and bottom.

A cap or crown structure is used on the top and bottom of the TF coil set, and additional structure is added around and outside the TF coil outer legs. The modified shape of the coils and the tailored structural elements provide room to extract the removable power core sector. Figure 1 shows the modified coil shape that is reduced in overall height and increased in the horizontal direction. An added benefit is a reduced field requirement for the poloidal field (PF) coils which reduces the size, current, and cost as compared to a constant-tension, D-shaped TF coil set that would provide similar access. The PF coil set was also designed without coils in the path of the maintenance port so that sectors could be removed.

Figure 1. Section view of power core components

One of the design features [2] is the grading of the in-vessel components so that the useful end-of-life in each zone is roughly similar and all material in a zone may be removed at the same time. It is especially beneficial to have the same lifetime for the divertor as the first wall and blanket, which will allow all these components (within a sector) to be removed and refurbished simultaneously after 2.5 full power years (FPY).

Surrounding the innermost FWBD zone is an integral hardback support structure that also functions as a part of the blanket, tying all these components together within the power core, including during maintenance operations. The primary support for this sector module is from the bottom, with locking keys or pins on the inner wall and the upper region. This support structure will be removed with the FWBD zone but will be refitted with a new FWBD set of components and returned to service until it accumulates a service life of 7.5 FPY.

Further inboard are a high temperature shield and a low temperature shield, both of which are life-of-plant components with a lifetime of approximately 40 FPY. These components are not removed during normal maintenance operations. On the outboard side, it is more efficient to remove the FWBD, hardback, shielding, and vacuum vessel/door assembly as a single piece. This will allow all the blanket and shielding plumbing to remain with the sector module as it is being removed. The parting line for the removable sector continues just above (or below) the upper (or lower) FWBD structure and horizontally out beyond the vacuum vessel. The vacuum vessel is adjacent to the low temperature shield to make the assembly as compact as possible and minimize the mating structure. The maintenance port is a separate vacuum zone to help isolate any coolant or tritium leaks. All plumbing connections are made in this zone to ease maintenance operations.

The previous discussions dealt with the periodic, scheduled maintenance of the power core components. The design must also address the unlikely failure of a nominal life-of-plant component such as a TF or a PF coil failure. The cryostat has a removable dome for access to the upper or central PF coils for replacement. If a TF coil fails, the upper PF coils may be removed and the life-of-plant core components such as the shielding, vacuum vessel, and TF cap will be dismantled and removed. Then the failed TF coil will be vertically removed and replaced. If a lower PF coil would fail, all of the above would have to be removed or the coil would have to be remotely rewound in place. This lower PF coil failure was considered very unlikely, but the consequence is so onerous that the project elected to install spare coils below the operating coils. The much simplified replacement of the lower PF coils would only involve remotely cutting out the failed unit, raising the spare into place, and making the electrical and cryogenic connections.


The viability of this sector module maintenance approach can only be validated with 3-D CAD models of the complete sector being removed through a specially designed vacuum vessel maintenance port. Figure 2 is a plan view at the midplane of the power core that depicts several of the critical features and requirements of this design. This view illustrates the boundary of the sector removed. The inboard portion of the removable module includes the first wall/blanket and the replaceable shield/hardback. The module extends radially outward to the outermost portion of the blanket/reflector/hardback region so that all the shorter lived components can be fully replaced. A small wedge of shielding and vacuum vessel structure will remain in the core region when the sector is removed. To improve maintenance times and ease of access, the outboard shields are withdrawn as a part of the sector. In the hot cell, all parts with remaining useful life are inspected for reuse and then mated with new replaceable components and returned to service at the next maintenance action.

Figure 2. Midplane section view of power core

The vacuum door immediately outside the shields is sealed with a welded joint. Both welded and mechanically-sealed door designs were examined, but at this time a welded seal appears simpler and can be cut and rewelded quickly using modern welding techniques.

Figure 3 is a solid model to help visualize the configuration of the removable sector. This figure also illustrates that the divertor region is an integral part of the removable sector.

Figure 3. Removable sector

Figure 4 is a cutaway of the entire power core through one of the maintenance ports. One of the sectors is shown partially withdrawn out of a port. The mover would attach to the sector, slightly raise the sector to support the load, and withdraw it into an individually-sealed transporter for removal to the hot cell for refurbishment.

Figure 4. Sector being withdrawn from Power Core (shields and building not show for clarity)


In addition to the in-vessel sector design, several other important system requirements influence the maintenance scheme and the power core building. These include:

Several options to access the ports and transport the sectors to the hot cell were assessed.

  1. One of the simplest was to remove a bare sector out onto a pad, with a crane assist, and transport it to the hot cell. But this freely distributes any radioactive particulate to all surfaces in the power core building, and therefore was rejected.
  2. A slight improvement was to use a close-fitting tunnel. The surface area of the tunnel is still sizable in this case and would require either periodic decontamination or full isolation.
  3. Another option was to use some form of moving enclosure, such as a transporter or flask.

A moving enclosure concept was adopted for this design. The transporter could be small enough to accommodate a single sector or large enough to accommodate several ports and sectors. The eventual choice will depend on the time necessary to transport the transporter or flask, access the ports, withdraw the sector, and install a new sector. The transporter for a single sector provides greater flexibility, but it must make two trips: once empty to the port to remove a sector back to the hot cell, and then again to return with a new one. Then it can move over to a new location and begin again. Single transporters are less expensive and can accommodate random ports, but also require a slightly larger maintenance area for operations. Multiple-sector flasks will be more expensive but will accommodate several adjacent ports so fewer trips will be required. Parking space for new and used sectors must be provided within the flask as well as space for sector movers. The final choice between a single vs. multiple sector approach will depend on more detailed trade studies on the time to transport the transporter orflask, access the ports, withdraw the sector, and install a new sector. For this paper, the single transporter is illustrated to show the principles of maintenance.

Figure 5 is a plan view through the horizontal maintenance ports and maintenance corridor illustrating the power core elements surrounded by a primary bioshield with doors for each maintenance port opening. Remotely operated or autonomous shield movers will remove the shielding doors and place them alongside the port openings. Isolation doors at the cryostat radius will be closed to prevent dispersal of contaminants to the maintenance corridor. These doors will be opened only when the transporter is docked.

Figure 5. Transporters and movers conducting periodic maintenance

The transporters could be large enough to house only the removable sectors if the sectors were self-moving. But this was not considered practical because the mechanisms would be subject to the power core environment over its useful life. Instead, a separate remote handling mover will be used. Options to move sectors into the transporter unit include a rail system, roller bearings, or a cantilevered lifting device. Each has its own problems and needs to be investigated in more detail.

Figure 5 also shows a transporter docked to a port with the sector mover beginning to withdraw a sector into the transporter. Another transporter is en route to the airlock. Meanwhile another transporter is returning to the hot cell with a used sector. This illustrates that there is sufficient clearance to maneuver around docked transporters and conduct operations at several ports.


The ARIES-RS conceptual design presents a highly maintainable tokamak fusion power plant that should be capable of efficient periodic maintenance of the irradiated power core components. This design represents the initial step toward an integrated and maintainable power core and maintenance system.


1. F. Najmabadi and the ARIES Team, "The Starlite Project: The Mission of the Fusion Demo," 16th IEEE Symposium on Fusion Engineering, Sept 30.-Oct. 5, 1995, Champaign IL.

2. M. S. Tillack and the ARIES Team, "Engineering Overview of the ARIES-RS Tokamak Power Plant," these proceedings.


* Work supported by the US Department of Energy under contract DE-AC03-95ER-54299 and UCSD Purchase Order 10087872.

** Institutions participating in the ARIES Team, in addition to McDonnell Douglas Corporation, include Argonne National Laboratory, General Atomics, Los Alamos National Laboratory, Massachusetts Institute of Technology, Princeton Plasma Physics Laboratory, Raytheon Engineers and Constructors, Rensselaer Polytechnic Institute, University of California-San Diego,and the University of Wisconsin-Madison.