Performance assessment of primary frequency control responses for thermal power generation units using system identification techniques. (September 2018)

• Record Type: Journal Article
• Authors: Wang, Jiandong; Su, Jianjun; Zhao, Yan; Pang, Xiangkun; Li, Jun; Bi, Zhenfu
• Is Part Of: International journal of electrical power & energy systems. Volume 100(2018)
• Page Start: 81
• Page End: 90
• Publication Date: 2018-09
• Subjects: Primary frequency control -- Performance assessment -- System identification -- Thermal power generation units; Electrical engineering -- Periodicals; Electric power systems -- Periodicals; Électrotechnique -- Périodiques; Réseaux électriques (Énergie) -- Périodiques; Electric power systems; Electrical engineering; Periodicals; 621.3
• DOI: 10.1016/j.ijepes.2018.02.036
• Languages: English
• ISSNs: 0142-0615

Abstract

Highlights: A novel method is proposed to estimate performance metrics of PFC responses. All data points in PFC responses are exploited in identifying dynamic models. The proposed method is applicable to arbitrary-types of PFC responses. Effects from main steam pressures are separated in performance metric estimation. Industrial examples from several thermal power generation units are provided. Abstract: Primary frequency control (PFC) responses for thermal power generation units are concerned about whether changes of turbine rotating speeds lead to proper responses of generator active powers. A new method is proposed to estimate performance metrics for PFC responses via system identification techniques. Dynamic models are identified from data samples of the turbine rotating speed and the main steam pressure as two inputs, and the generator active power as the output. Hypothesis tests are formulated to evaluate the qualities of identified models. If model qualities are satisfactory and the contribution from the turbine rotating speed to the generator active power is significant, then performance metrics are estimated from a noise-free unit step response of the identified model. The proposed method removes three major limitations of the most widely-used method in contemporary industrial practices, namely, (i) performance metrics are obtained from all data samples of PFC responses, instead of some specific data samples being sensitive to noises; (ii) the proposed method is applicable to arbitrary-type PFC responses, not limited to step changes of the turbine rotating speed; (iii) noticeable effects of the main steam pressure are separated to avoid erroneous performance assessment results. Examples from thermal power generation units are provided to illustrate the effectiveness of the proposed method.