A new method for simultaneous optimizing of wind farm's network layout and cable cross-sections by MILP optimization

- Simultaneous optimization of wind farm's network layout and cable cross-sections.
- Proposed method brings substantial cost savings compared with other methods.
- Usefulness and effectiveness of the MILP optimization was demonstrated.
- Electrical losses need to be taken into account in the design process.
- An absolute optimal solution (GAP = 0) is received in a reasonable computation time.

Internal electrical networks of large wind farms constitute complex and dispersed grid structures. Wind turbines are scattered over vast areas and the total length of cabling infrastructure might reach several dozens of kilometers. Outlays related to cable laying significantly contribute to the entire project budget. Therefore the design process should minimize these expenses considering also operation and maintenance costs calculated over the project lifetime on condition of fulfilment of all technical requirements. An analysis presented in this paper demonstrates that an independent optimization of the twofold problem dealing both with investment and operation costs does not result in the cheapest solution. The analysis confirmed also reliability and effectiveness of application of Mixed Integer Linear Programming method (MILP) to solve this kind of optimization problem. The paper shows that the developed integrated optimization algorithm is efficient and delivers an absolute optimal solution (GAP = 0) in a reasonable computation time. The results obtained for a real wind farm project confirm that the optimal design of a wind farm network can't be determined a priori and the final outcome strongly depends on the configuration of wind turbines (e.g., number of feeders, number of turbines connected to a single feeder, etc.) and technical parameters of cables. Spread over time, discounted costs of energy losses are an integral part of the objective function. The study proves that cost of energy losses impacts on the overall financial results and shouldn't be neglected. The related expenses are roughly at the same level as expenditure linked to cable laying and they heavily influence the final design of the internal network. The results show the possibility of practical use of the proposed algorithm in the wind farm design process.