The choice of location for new wind, solar, or gas power plants currently depends primarily on available land and the generation costs at the respective site – while the costs of grid expansion and congestion management are not taken into account. Several reform proposals, in particular the current draft legislation on the “grid package” (Netzpaket), aim to change this. The analysis compares seven currently discussed or already implemented instruments for the spatial steering of investments in electricity generation facilities: redispatch reservation (Redispatch-Vorbehalt), construction cost subsidies (Baukostenzuschüsse), flexible connection agreements, bidding zone splitting, dynamic grid fees, as well as spatially differentiated remuneration under the Renewable Energy Sources Act (EEG) and within capacity mechanisms. The instruments are first considered individually in terms of their effect on the profitability of generator investments at sites with high and low grid costs. On this basis, the analysis evaluates their efficiency effects (locational efficiency, dispatch efficiency, and investment certainty) as well as their distributional effects on generators, grid operators, consumers, and the state. Possible combinations of the instruments and their interactions are then discussed.
None of the instruments steer investment without drawbacks: all of them can improve locational efficiency, but predominantly with side effects on dispatch efficiency or investment certainty. Bidding zone splitting, along with spatial steering via the EEG or capacity mechanisms, initially appears relatively efficient, but entails considerable disadvantages in implementation. Redispatch reservation, flexible grid connection agreements, and dynamic grid fees tend to worsen dispatch efficiency and investment certainty, and these disadvantages can only be partially remedied. A combination of construction cost subsidies and feed-in limitations, by contrast, proves promising: if construction cost subsidies were levied only on a firm partial connection capacity, and an optional, dynamically limited feed-in were additionally enabled, both grid congestion and grid expansion costs could be internalized, while disadvantages with regard to dispatch efficiency and investment certainty could be reduced.
