Flexibility from distributed deferrable loads presents an enormous potential to provide fast ramping resources that are necessary to vastly integrate renewable energy resources. In this paper, we study aggregation, characterization, and scheduling of a collection of deferrable loads to facilitate integrating renewable generation into the power system. A generation profile is called exactly adequate if there exists a scheduling policy that could allocate the power to satisfy the energy requirements of all deferrable loads without surplus or deficit. We provide sufficient and/or necessary characterizations on the adequacy of allocated generation profiles. Moreover, we propose a novel scheduling algorithm to service deferrable loads. Heuristic algorithms such as Earliest Deadline First (EDF) and Least Laxity First (LLF) policies are used in numerical experiments to compare with the proposed algorithm. Extensive simulations show that our scheduling policy generally can fulfill the energy requirements of all loads without surplus or deficit for exactly adequate generation profiles, while the EDF and LLF policies cannot meet this objective in most cases.