As the technology readiness level of ocean energy devices moves from single units to small arrays, evidence from related industries suggests the need for integrated array design tools. Those tools implement all the functionality required for assessing the economic performance, technical reliability and environmental impact of the diverse subsystems to be accounted for in the design chain. These include device hydrodynamics, design of the electrical connection to shore, moorings and foundations design, as well as logistic requirements for the installation and maintenance of the farm. This paper presents a critical discussion of approaches for optimising the solutions of such a tool when the target of the problem is to achieve the lowest Levelised Cost of Energy (LCOE) solution. Firstly, the problem is formally described, pointing out the complexity of the architecture of the models, the intervening variables and the constraints and boundary conditions of the solution space. Then, approaches for identification of lowest LCOE solutions are examined, starting from simplified one-parameter sensitivity analyses-such as, for example, interdistance among devices, and distance to shore-to more complex heuristic approaches. All the above considerations will be illustrated through examples pertinent to the ocean energy sector.