Due to the growing share of variable renewable energy sources, system integration and the contribution of mature renewable technologies to security of supply is becoming increasingly important. This implies that the overall infeed from renewables should be as high as possible at times of high electrical load. Hence, it is questioned whether variable renewables can contribute to the improvement of system reliability. This paper investigates to what extent a geographical diversification of wind onshore farms can support generation adequacy. Based on meteorological data, synthetic time series of wind infeed are simulated and geographical allocation of wind turbines is optimized. A probabilistic convolution approach is proposed to determine required firm capacity nececsary to maintain security of supply. The findings indicate that changing the allocation of wind turbines allows saving up to 2 GW of required firm capacity. In case of wind turbines with storm control, the wind feed-in during the system peak load can be increased by about 0.5 GW. From an economic perspective a trade-off between maximizing the wind yield and contribution of wind to generation adequacy of 0.8 bn(sic) is found. Further possible cost savings regarding balancing and transmission should be incorporated in the context of renewables policy design.