Operational intelligence in electric power systems is focused in a small number of control rooms that coordinate their actions. A clear division of responsibility and a command hierarchy organize system operation. With multi-agent based control systems, this control paradigm may be shifted to a more decentralized open-access collaboration control paradigm. This shift cannot happen at once, but must fit also with current operation principles. In order to establish a scalable and transparent system control architecture, organizing principles have to be identified that allow for a smooth transition. This paper presents a concept for the representation and organization of control- and resource-allocation, enabling computational reasoning and system awareness. The principles are discussed with respect to a recently proposed Subgrid operation concept.