Accurate representation of frequency-dependent cable parameters is essential for evaluating power losses and voltage regulation in medium-voltage wind-farm collection networks. This paper develops a MATLAB-based electromagnetic framework using frequency-dependent series-impedance (Z) and shunt-admittance (Y) matrices for underground cables, including conductor skin and proximity effects and earth-return impedance. Four earth-return formulations—Saad–Gaba–Giroux, Wedepohl series, Wedepohl, and infinite-earth—are implemented and benchmarked against RSCAD over the frequency range from 10 Hz to 1 MHz. The validated model is applied to a wind-farm collection system in the Isthmus of Tehuantepec, Oaxaca, Mexico, comprising five 1.5 MW wind turbines connected at 13.8 kV through two radial underground circuits. The results show that the Saad–Gaba–Giroux and Wedepohl-series formulations provide the closest agreement with the RSCAD reference. At a power factor of 0.9, the conventional constant core-resistance model underestimates the modeled active-power loss by up to 60% under the evaluated operating conditions. The frequency-dependent model also predicts voltage drops of 0.68% and 1.55% for Circuits 1 and 2, respectively, whereas the simplified DC representation predicts values below 0.15%. A component-level analysis identifies the core conductor as the dominant contributor to active-power dissipation, while the earth-return path has a significant influence on the equivalent phase resistance and voltage-drop behavior. These results demonstrate that frequency-dependent cable modeling can provide a more comprehensive basis for technical assessment and techno-economic evaluation of medium-voltage wind-farm collection systems.

Accurate representation of frequency-dependent cable parameters is essential for evaluating power losses and series-voltage effects in medium-voltage wind-farm collection networks. This paper develops a MATLAB-based electromagnetic framework that derives frequency-dependent series-impedance (Z) and shunt-admittance (Y) matrices for underground cables, including conductor skin and proximity effects and earth-return impedance. Four earth-return formulations—Saad–Gaba–Giroux, Wedepohl series, Wedepohl, and infinite-earth—are implemented and benchmarked against RSCAD from 10 Hz to 1 MHz. The AC series parameters evaluated at 60 Hz are applied to a 7.5 MW wind farm in the Isthmus of Tehuantepec, Oaxaca, Mexico, comprising five 1.5 MW turbines connected at 13.8 kV through two radial underground circuits. The Saad–Gaba–Giroux and Wedepohl-series formulations provide the closest agreement with the RSCAD reference. At a power factor of 0.9, the simplified DC core-resistance baseline underestimates active-power losses by 34.35% in Circuit 1 and 42.83% in Circuit 2 relative to the 60 Hz AC series-resistance model. At rated turbine output and the same power factor, the accumulated AC series-voltage indicators at the farthest turbine are 0.17% and 0.38%, compared with 0.11% and 0.20% using core-only DC resistance. These indicators do not represent terminal-voltage magnitudes, which require an AC network power-flow calculation. A component-level analysis for Circuit 1 identifies the core conductor as the dominant contributor to active-power dissipation, followed by the earth-return path and metallic sheath. An annual economic evaluation yields similar project-level financial indicators for the AC and DC representations, while a parametric resistance sensitivity analysis shows decreasing profitability and longer investment recovery periods as AC cable resistance increases. The frequency sweep validates the cable-parameter formulations; their application to electromagnetic-transient simulations remains future work.