The Black Sea is a nearly enclosed basin with a highly productive northwestern shelf that experiences seasonal bottom hypoxia. Shelf–basin exchanges are controlled by the Rim Current and eddy activity, while the basin interior is strongly shaped by the Cold Intermediate Layer (CIL), whose formation and persistence influence stratification, ventilation, and biogeochemical cycling. Capturing these coupled physical–biological interactions in multi-decadal simulations is challenging because biogeochemical models are computationally expensive—the BAMHBI biogeochemistry model adds 28 advected tracers and roughly quadruples computational cost compared to physics-only runs. This has motivated long integrations at coarse horizontal resolution (~15 km), but this setup poorly represents coastal geometry, coarsely resolves the Rim Current, and misses finer-scale circulation features that likely matter for cross-shelf transport and oxygen dynamics. We're investigating whether moving to higher resolution affects the physical processes that control oxygen budgets and hypoxia on the shelf. We compare three NEMO–BAMHBI configurations over 2000–2024: (i) 15 km with 59 vertical levels, (ii) 5 km with 59 levels, and (iii) 5 km with 75 levels, with vertical refinement across the CIL depth range (50–200 m). We're looking at how resolution affects oxygen budgets on the shelf (including hypoxia extent and duration), transport of oxygen and organic matter from shelf to basin, impacts on primary production and chlorophyll distributions, and the formation and vertical structure of the CIL. While 5 km resolution doesn't explicitly capture submesoscale processes, improving the representation of mesoscale circulation and stratification should provide a more realistic foundation for understanding cross-shelf transport and hypoxia dynamics. This work aims to establish a computationally feasible yet sufficiently resolved configuration for future multi-decadal simulations in the Black Sea.