In-silico design of polymeric biomaterials requires molecular dynamics (MD) simulations that retain essential atomistic/molecular details (e.g., explicit water around the biofunctional macromolecule) while simultaneously achieving large length and time scales pertinent to macroscale function. Such large-scale atomistically detailed macromolecular MD simulations with explicit solvent representation are computationally expensive. One way to overcome this limitation is to use an adaptive resolution scheme (AdResS) in which the explicit solvent molecules dynamically adopt either atomistic or coarse-grained resolution depending on their location (e.g., near or far from the macromolecule) in the system. In this study we present the feasibility and the limitations of AdResS methodology for studying polyethylene glycol (PEG) in adaptive resolution water, for varying PEG length and architecture. We first validate the AdResS methodology for such systems, by comparing PEG and solvent structure with that from all-atom simulations. We elucidate the role of the atomistic zone size and the need for calculating thermodynamic force correction within this AdResS approach to correctly reproduce the structure of PEG and water. Lastly, by varying the PEG length and architecture, we study the hydration of PEG, and the effect of PEG architectures on the structural properties of water. Changing the architecture of PEG from linear to multiarm star, we observe reduction in the solvent accessible surface area of the PEG, and an increase in the order of water molecules in the hydration shells.