One of the biggest concerns associated with integrating a large amount of renewable energy into the power grid is the ability to handle large ramps in the renewable power output. For the sake of system reliability and economics, it is essential for power system operators to better understand the ramping features of renewables, load, and netload. In this paper, an optimized swinging door algorithm (OpSDA) is adopted and extended to accurately and efficiently detect ramping events. For wind power ramps detection, a process of merging "bumps" (that have a different changing direction) into adjacent ramping segments is integrated to improve the performance of the OpSDA method. For solar ramps detection, ramping events that occur in both clear-sky and measured (or forecasted) solar power are removed to account for the diurnal pattern of solar generation. Ramping features are extracted and extensively compared between load and netload under different renewable penetration levels (i.e., 9.77%, 15.85%, and 51.38%). Comparison results show that: (i) netload ramp events with shorter durations and smaller magnitudes occur more frequently when renewable penetration level increases, and the total number of ramping events also increases; and (ii) different ramping characteristics are observed in load and netload even at a low renewable penetration level. (C) 2017 Elsevier Ltd. All rights reserved.