Maybe just run self-consumption for a year and see how the bills turn out?In fact I don't know if prioritizing solar for self-consumption wouldn't knock me over the threshold.
If you're a net zero kWh importer/exporter, then the only way you have a bill in dollars is NBCs and if you are net importer during peak. The more you are a net kWh exporter, the more you can afford to be a net importer during peak, as it will be offset by all your net exports during off-peak. So minimizing NBCs may be more important cost-wise than avoiding any grid draws during peak.
You could also try TBC with all your sell rates set to 7 cents/kWh, and your buy costs set to what you'd pay if you aren't a net dollar exporter. That should incentivize the control algorithm to self-consume rather than sell back, and to prefer off-peak grid draw to peak grid draw. It won't incentivize it to time shift production to export at peak, but seeing as how you are a net kWh exporter, you don't also need to be a net kWh exporter during peak to avoid having a bill in dollars.
Any experiment like the above should be run at least a week to give the algorithm a chance to stabilize, probably more like a month.