We present a new systematic configurational sampling algorithm for investigating the potential energy surface of hydrated atmospheric molecular clusters. The algorithm is based on creating a Fibonacci sphere around each atom in the cluster and adding water molecules to each point in nine different orientations. For the sampling of water molecules to existing hydrogen bonds, the cluster is displaced along the hydrogen bond, and a water molecule is placed in between in three different orientations. Generated redundant structures are eliminated based on minimizing the root-mean-square distance of different conformers. Initially, the clusters are sampled using the semiempirical PM6 method and subsequently using density functional theory (M06-2X and ωB97X-D) with the 6-31++G(d,p) basis set. Applying the developed algorithm, we study the hydration of sulfuric acid with up to 15 water molecules. We find that the addition of the first four water molecules "saturate" the sulfuric acid molecule and that they are more thermodynamically favorable than the addition of water molecules 5-15. Using the large generated set of conformers, we assess the performance of approximate methods (ωB97X-D, M06-2X, PW91, and PW6B95-D3) in calculating the binding energies and assigning the global minimum conformation compared to high level CCSD(T)-F12a/VDZ-F12 reference calculations. The tested DFT functionals systematically overestimate the binding energies compared to coupled cluster calculations, and we find that this deficiency can be corrected by a simple scaling factor.