Identifying torsional modal parameters of large turbine generators based on the supplementary-excitation-signal-injection test

• The new approach is based on the supplementary-excitation-signal-injection test.
• It can identify frequencies and mechanic damping of all torsional modes concerned.
• A solid theoretical foundation is established upon the dynamic shaft model.
• A systematic test and identification procedure is developed.
• The approach is applied in practical systems and validated through field test.

Torsional parameters, especially modal frequencies and mechanical damping, of large turbine generators play critical roles in evaluating and solving the subsynchronous resonance or oscillation (SSR/SSO) problem. To accurately identify these torsional parameters, this paper proposes a systematic approach based on the supplementary-excitation-signal-injection test. The identification process is fulfilled via three steps, i.e., (i) accurate detection of modal frequencies by stimulating controllable torsional vibration with the injection of supplementary modal signals into the excitation system; (ii) online identification of total modal damping using modal filtration, improved discrete Fourier transform (DFT) and least-square fitting technique; (iii) separating electrical damping from the identified modal damping to get the pure mechanical modal damping. The proposed approach was verified through digital simulation and then was applied to the torsional-parameter identification of four practical turbine-generators in Shangdu Power Plant. The results demonstrated that, with the proposed approach, the modal frequencies and mechanical damping can be obtained accurately, intactly and without interfering with the normal operation of on-grid generators.