There is an increased use of problem representations (i.e. superstructures in synthesis problems; networks in route problems; graphs; ordered graphs in various systems representations) following on significant advances in optimization technologies that hold capabilities to solve, robustly, large-scale problems. In an attempt to systematically tackle disparate domains and build high-throughput functions, the paper contributes with a semantically-enabled approach systematized and engineered by ontologies. The aim is to develop an intelligent environment with capabilities to build and scale-up system representations, automatically. The work is demonstrated on problems akin to biorenewables and biorefineries; an identical approach is possible to the general problem. Using relations and rules defined among entities, semantics are deployed to model and expand domains (biorefinery pathways) whereas enabling extracting and creating knowledge. The repository, already on a web-based platform and available as open-source, essentially upgrades conventional representations with capabilities to share (import/export) and integrate its content externally. (C) 2018 The Authors. Published by Elsevier Ltd.