In previous publications [1,2], it was rationalized that a large vertical potshell deformation may have a negative impact on the operations of very high amperage cells.
The MHD-Valdis non-linear Magneto-Hydro-Dynamic model was therefore extended to take into account the displacement of the potshell. The MHD cell stability behavior of a 500 kA cell with a 17.3 meters long potshell was then studied.
Introduction
The work presented in this paper is part of a longer term collaboration efforts by the authors to investigate if there is a technical limit to the size of an aluminum electrolysis cell that can be designed, built and successfully operated.
The first and currently most popular argument is that there is a size limit dictated by the heat dissipation requirement hence the need for the AP50 technology to use an active heat exchanger system on the potshell in order to enhance the heat dissipation.
This issue was addressed in [3]. The demonstration is quite simple to make: because there is a size restriction on the cell width, having a bigger cell means having a longer cell. Already, length to width aspect ratio of modern high amperage cells is quite high so they proportionally do not dissipate much heat by their end walls. Very simple calculations can be made to demonstrate that a 300 kA cell operating at 15 kWh/kg would need higher heat fluxes to dissipate its internal heat than a 740 kA cell operating at 13.5 kWh/kg.
The second argument is that there is a size limit dictated by the MHD cell stability requirement. This issue was addressed in [1]. It is known that the modern high amperage cell operating at around 300 to 350 kA can comfortably be operated with asymmetric busbar network compensating for the effect of the return line located about 60 meters away. Furthermore, the Bz minimization requirement imposes that the great majority of the positive side busbars must go around the cell.
At 500 kA, the cell aspect ratio is such that it becomes both impractical and quite expensive to continue to run most of the positive side busbars around the cell. In order to avoid to have to do that, a compensation busbars network like the one presented in [4] is required.
While until further notice, it seems true that at around 500 kA, a compensation busbars network is required in order to address both the cell stability problem and the busbar cost minimization problem, it was claimed in [1] that it is possible to perfectly magnetically compensate cells of any length and hence of any amperage as results presented for a 500 kA and a 740 kA cell clearly demonstrate.
The third and last argument is that there is a size limit dictated by the cell mechanical design requirements. This argument is not often even raised, but as discussed in [2], it may well be the most serious argument.
It used to be that due to carbon lining swelling, the potshell badly deformed laterally. But modern high amperage cells using 100% graphitic or graphitized cathode blocks and strong orthotropic potshell do not deform much laterally due to carbon swelling they rather deform vertically due to their own thermal loading [1,2].
It can be observed (but it has not yet been formally presented in a paper) that this vertical deformation can have a negative impact on the cell operations. The aim of the present paper is to take the first step in trying to take into account impact of the vertical potshell deformation on the cell predicted operational behavior.
Fully coupled aspect of the problem
To start with, the vertical potshell deformation problem is a fully coupled thermo-electro-mechanical problem. The potshell deforms due to its thermal load and this thermal load is generated by the cell heat dissipation characteristics [2].