background image
The task at this stage is to enter all that information into Dyna/Marc and to ask: at what amperage
does that cell design need to operate in order to be in perfect thermal balance at a typical cell
superheat? As can be seen in Fig. 10, the Dyna/Marc answer to that question is 600 kA.
Verification of the thermal balance at 600 kA using the ANSYS® based TE models: Dyna/Marc
is the perfect tool to get a quick answer to difficult questions like the one asked above; but this
answer cannot be accepted as final. It is safe standard practice to double check this Dyna/Marc
prediction using the more accurate ANSYS® finite element based TE models. The model can be
either a full cell slice model or, as in this current study, can combine models of separate half anode
and cathode side slice.
The half anode model predicts that 48 2.0 m long anodes in a cell operated at 600 kA will have an
average anode drop of 318 mV, and they will dissipate 449 kW with a 10 cm thick cover. The
cathode side slice model predicts that if operated at 600 kA and 7°C of superheat, then the cathode
drop would be 104 mV and the cathode would dissipate 676 kW, while maintaining a comfortable
ledge profile.
Since the internal heat at 600 kA, (which corresponds to an anodic current density of 0.94 A/cm2
and 3.5 cm ACD), according to Dyna/Marc will be 1,140 kW, the cell should be in perfect thermal
balance quite close to those assumed operating conditions. Note that Dyna/Marc also predicts 96.4%
current efficiency (CE) and 4.29 V, which corresponds to a energy consumption of 13.3 kWh/kg Al.
Verification of the the MHD stability at 600 kA using MHD-Valdis: The final verification to
make is that the cell will still be stable when operated at 600 kA without any modification of the
busbar. The answer may depend on the type of busbar design selected for the base case 500 kA cell
technology. An asymmetric busbar designed to auto-compensate a 500 kA return line will not be
able to perfectly compensate a 600 kA return line, and that will for sure reduce the cell stability.
Yet a very good busbar design is able to accommodate a lot of amperage creep, as we learned from
the evolution of the AP30 cell technology over the last 20 years. The busbar was initially designed
to operate at 280 kA, and yet that same busbar now supports cell operation at 360-380 kA, still
without major impact on the cell stability. Clearly there is some built-in robustness in a good busbar
design!
Nevertheless, a busbar design incorporating independent compensation busbars is more flexible as it
allows separately adjustment of the current running in the compensation loop(s). Suitable
adjustments can perfectly compensate for increased return line current. Fig. 11 shows results for the
third busbar design, this time with a cell operating at 600 kA instead of 500 kA . This demonstrates
that, despite the amperage increase the Bz is still more or less the same once the amperages in the
compensation loops are adequately readjusted.
Note that the ACD has not been readjusted between the 500 kA and the 600 kA runs of the MHD
cell stability analysis. This is because we assume that it is the use of slotted anodes that leads to a
smaller calculated ACD and so changes the internal heat generation. The physical ACD that matters
in terms of cell stability we assume remains the same in the 500 kA cell design (using the old
conventional unslotted anodes) as in the 600 kA cell design (using slotted anodes).
Conclusions