We did a rough calculation and found that with realistic self-consumption rates (achievable without absurd effort to increase the SCR, i.e., 20-30%) at current electricity prices, full feed-in is generally more advantageous (for 10 kWp, 8.11 vs 12.87 cents / kWp). This changes if one assumes a moderate increase in electricity prices (e.g., like the last 20 years, about 4% per year) or manages to raise the self-consumption rate somehow. However, there is also an argument against increasing electricity prices: the energy transition. If it succeeds, electricity will potentially not get more expensive. Because generation would become cheaper, only grid fees would rise. CO2 as well, but the CO2 costs cannot be passed on to renewable energies. So wait until 2025. If CDU wins next year and it looks like a grand coalition, better to rely on self-consumption. If the Greens somehow manage again, then rather full feed-in – provided it is absolutely necessary to make a business case out of it. If you have many differently oriented roof surfaces, you could also register multiple systems. For example, an east-west system as a self-consumption system and the rest (especially south) as full feed-in.