Notwithstanding its variability and limited controllability, wind power is expected to contribute strongly to electricity generation from renewable energy sources in the coming decades. Treating wind power as non-dispatchable by subtracting its output from the original load profile, results in a net load profile, which must be covered by conventional power generation. The screening curve methodology is a first approximation to find the optimal generation technology mix, based on relative cost levels. However, increased variability of the net load profile, due to wind power generation, strongly influences system operation. Therefore a static linear programming investment model is developed to determine the optimal technology mix. This alternative methodology shows a reduced capacity of inflexible generation after including operational constraints to properly account for net load variability. In order to illustrate this methodology, an example is set up, showing the sensitivity with respect to ramp rates of conventional generation, transmission interconnection and energy storage. The comparison of those different sources of system flexibility suggests that energy storage facilities better facilitate the integration of wind power generation. (c) 2010 Elsevier Ltd. All rights reserved.