Applications such as the flow around a wind-farm are typically associated with high complexity due to their size, further exaggerated when considering the large number of different environmental conditions that need to be studied. High-fidelity methods, despite their maturity and ability to provide solutions to such problems, lead to prohibitively large systems. An alternative is the use of reduced order models (ROMs), which have proven their ability to accurately, yet efficiently approximate the complex underlying system. Even then, the construction of a sufficiently parametrised ROM for such large problems can prove computationally taxing. As a solution, this work explores the viability of a ROM component that would replace its high fidelity counterpart in the original problem, resulting in a coupled chain-ROM strategy, where the output of upstream reduced components will act as input for downstream ones. The ROM component accepts highly parametrised inputs covering a wide range of inflow conditions, while special care is required in the reduction of the parametrisation complexity and in the handling of the interface between ROM components and nested boundaries. Moreover, the proposed ROM component uses a hybrid scheme combining data-driven and projection-based techniques. Velocity and pressure are approximated using the proper orthogonal decomposition (POD) with Galerkin projection, while eddy viscosity is approximated using POD plus interpolation. The parametrised flow around a series of wings is used to demonstrate the potential of the proposed methodology.