Technical Report NTB 09-05

Critical Review of Welding Technology for Canisters for Disposal of Spent Fuel and High Level Waste

Nagra is the Swiss national cooperative for the disposal of radioactive waste and is responsible for final disposal of all types of radioactive waste produced in Switzerland. As part of Nagra's long term disposal strategy, plans must be developed for two repositories, one for spent fuel (SF), vitrified high-level waste (HLW) and long-lived intermediate level waste and one for low and intermediate level waste. Within the next 10 years, Nagra plans to apply for general licences for these repositories. In the application, documentation will be required showing that long-term safety can be ensured and that factors for the construction, operation, and closure of the facility have been considered. Nagra has commissioned TWI to carry out a critical review of welding technologies for the sealing of HLW and SF canisters made of carbon steel, one of the preferred materials under consideration. The information in this report will be used in conjunction with a material selection report already produced. This report is intended as a preliminary step to provide input to developing design concepts for the SF and HLW canisters.


The objective of this report was to carry out a critical review of all available welding technologies for the application of sealing carbon steel canisters for SF and HLW to be disposed of in a repository. The review discusses the following key variables:

  • Suitability of techniques for thickness of weld required.
  • Suitability for remote operation, maintenance and set-up.
  • Advantages and disadvantages of each welding process in terms of welding speed, weld quality, tolerances and cost.
  • Effect of welding process on parent materials properties including microstructure corrosion resistance, distortion and residual stress.
  • Potential post-weld treatments to reduce residual stress and enhance corrosion resistance.
  • Suitability of inspection techniques for the weld thickness required including remote operation and accuracy.
  • Impact of welding techniques on the canister design and material selection.
  • Critique of emerging technologies which may be suitable for the application in the future.

Work Carried Out

The review of potential welding technologies began with a feasibility review carried out by TWI experts in the relevant processes. Certain feasibility criteria were used to rule out processes clearly not suitable for the application. The next stage was to carry out research in the form of a literature review. This encompassed all remaining processes and was focused on looking for previous applications of the processes for the material and thickness suggested, and also safety critical applications such as applied in the nuclear and pressure vessel industry. Once the relevant information was gathered each process was reviewed individually by a TWI engineer with expertise and experience in the process. This information was used at a meeting to weigh up the advantages and disadvantages of each process and decide on which are likely candidate processes. These candidate processes were then reviewed further to include the likely metallurgical effects and potential inspection techniques.


  1. Based on available literature, TWI experience and the requirements set forth by Nagra, two processes offer the best solution:
    a) narrow gap tungsten inert gas (NG-GTAW) welding.
    b) electron beam welding (EBW).
  2. It is expected that the choice of the exact welding process and post-weld treatments required for the Nagra application can be made when more detailed acceptance criteria for the weld performance are established.
  3. Both NG-GTAW and EBW techniques are suitable and have a track record of welding within the thickness required of 60 - 150 mm.
  4. The travel speed and typical joint size have been discussed and it is likely that EB welding will be far more productive than NG-GTAW; however the joint is likely to be completed using NG-GTAW within 24 hours.
  5. Both processes are suitable for remote operation and were selected on the basis of reliability and repeatability without operator intervention. It is likely that more development will be necessary for the ancillary processes of NG-GTAW.
  6. All welding processes will have a deleterious affect on the parent material. Appropriate post-weld treatments will improve the material properties such that the risk of post-weld cracking will be mitigated but the properties will always be different from the parent material. Due to the rapid cooling rates the weld metal toughness in the as-welded condition for EB welds will be lower than for NG-GTAW.
  7. When welding thick carbon manganese (C-Mn) steel components, mitigation of residual stresses is likely to present a bigger challenge than distortion potential.
  8. A large weld deposit is thought to be more detrimental with respect to residual stresses than several smaller heat input passes and more work is required to understand this.
  9. The maximum residual stresses from EBW occur at the mid-thickness of thick section components and are tensile in all directions, with the maximum values located in the heat affected zone (HAZ) region. For NG-GTAW the maximum tensile residual stress is likely to be at a certain depth below the outer surface. This will depend on the constraints and geometry during welding.
  10. Further research is required for a thorough comparison on which method, NG-GTAW or EBW, generates the lowest residual stress magnitudes for the final design of the canister.
  11. Post-weld treatment is recommended to mitigate residual stresses. It is likely that a combination of the reviewed residual stress mitigation techniques will lead to the best technical solution.
  12. Post-weld heat treatment (PWHT) of C-Mn steels is typically carried out at 600 °C, for one hour per 25 mm of nominal weld thickness. This might not be suitable for the current application so other possibilities have been explored.
  13. Local heating redistributes the residual stresses in the region of the weld. This might be an economic option if EBW is used for sealing the canister since the equipment would already exist.
  14. Various surface treatment methods (shot, ultrasonic, hammer and laser peening) can be applied to modify the residual stresses near the external surface of the canister.
  15. Friction welding processes offer considerable opportunities for the mitigation of residual stresses. However, information is limited, and work would be needed to establish the suitability of these processes.
  16. More information is necessary on the mechanical properties required, before the impact of the design on the canister can be fully assessed. From a welding process viewpoint the integration of a self-locating spigot joint would be beneficial for controlling penetration and protecting the contents of the canister.
  17. Ultrasonic and radiographic inspection techniques are both appropriate for the non-destructive testing (NDT) of NG-GTAW and EB welds and it would be recommended that the processes are used in tandem to provide maximum assurance of the weld quality.
  18. The currently proposed material ASTM A516 Grade 70 is weldable with both processes; however, grades of steel with similar mechanical properties and corrosion resistance are available with improved weldability and lower impurity contents.
  19. Welding for the fabrication of the canister presents far less of a technological challenge than the sealant weld. It may be possible to produce a canister design that does not require fabrication welds.